Quick viewing(Text Mode)

The Legionella Type II Secretion Paradigm

The Legionella Type II Secretion Paradigm

REVIEW White and Cianciotto, Microbial Genomics 2019;5 DOI 10.1099/mgen.0.000273

Assessing the impact, genomics and evolution of type II across a large, medically important genus: the type II secretion paradigm

Richard C. White† and Nicholas P. Cianciotto*

Abstract The type II secretion system (T2SS) plays a major role in promoting bacterial survival in the environment and in human hosts. One of the best characterized T2SS is that of , the agent of Legionnaires’ disease. Secreting at least 25 proteins, including degradative enzymes, eukaryotic-like proteins and novel effectors, this T2SS contributes to the ability of L. pneumophila to grow at low temperatures, infect amoebal and hosts, damage lung tissue, evade the immune system, and undergo sliding motility. The genes encoding the T2SS are conserved across the genus Legionella, which includes 62 and >30 in addition to L. pneumophila. The vast majority of effectors associated with L. pneumophila are shared by a large number of Legionella species, hinting at a critical role for them in the ecology of Legionella as a whole. However, no other species has the same repertoire as L. pneumophila, with, as a general rule, phylogenetically more closely related species sharing similar sets of effectors. T2SS effectors that are involved in infection of a eukaryotic (s) are more prevalent throughout Legionella, indicating that they are under stronger selective pressure. The Legionella T2SS apparatus is closest to that of Aquicella (another parasite of amoebae), and a significant number of L. pneumophila effectors have their closest homologues in Aquicella. Thus, the T2SS of L. pneumophila probably originated within the order , with some of its effectors having arisen within thatAquicella -like progenitor, while other effectors derived from the amoebal host, mimivi- ruses, fungi and less closely related .

Legionella and that may represent novel species [15–18]. Of the confirmed pathogenesis Legionella species, which fall into three major phylogenetic clades, 32 are disease-causing, based on cultures obtained The genus Legionella was first recognized in the late 1970s, from symptomatic individuals or seroconversion. However, with the characterization of Legionella pneumophila as the aetiological agent of a form of now known as approximately 90 % of cases of Legionnaires’ disease in the Legionnaires’ disease [1, 2]. Within the Gammaproteobac- USA and Europe are caused by L. pneumophila [19]. Within teria, Legionella is the sole genus contained within the family aquatic systems, L. pneumophila and other legionellae Legionellaceae [3]. Members of this genus are Gram-negative primarily parasitize free-living . The host range of bacteria found ubiquitously in the environment in both Legionella species is exceptionally broad, as co-isolation and freshwater systems such as lakes and rivers, as well as man- co-culture experiments implicate permissive hosts within made aquatic systems [4–7]. There are at least 63 confirmed seven of the eight phyla within the protozoan kingdom, 12 of species of Legionella [8–14]. Additionally, there are a plethora 41 classes within those phyla, and 21 of 82 known orders [20]. of uncultured Legionella-like organisms in freshwater systems Some of the most abundant protozoa in nature, including

Received 25 March 2019; Accepted 08 May 2019; Published 05 June 2019 Author affiliations: 1Department of Microbiology and Immunology, Northwestern University Medical School, Chicago, IL 60611, USA. *Correspondence: Nicholas P. Cianciotto, n-​ cianciotto@​ northwestern.​ edu​ Keywords: Legionella; type II secretion; T2SS; Legionnaires' disease; Legionella pneumophila; Aquicella. Abbreviations: AP, assembly platform; BYE, buffered yeast extract; DUF, domain of unknown function; ER, ; HGT, ; HMM, hidden Markov model; IM, inner membrane; LCV, Legionella-containing ; ncRNA, non-coding RNA; OM, outer membrane; OMV, outer membrane vesicle; PPIase, peptidyl-proline cis/trans-isomerase; RefSeq, NCBI Reference Sequence Database; Sec, general secretory pathway; SG, serogroup; Tat, twin-arginine translocation pathway; T4P, type IV pilli; T2S, type II secretion; T2SS, type II secretion system; T4SS, type IV secretion system. †Present address: Department of Genomic Medicine, J. Craig Venter Institute 9605 Medical Center Dr., Rockville, MD 20850, USA. Data statement: Three supplementary tables are avaliable with the online version of this article. 000273 © 2019 The Authors This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1 White and Cianciotto, Microbial Genomics 2019;5

species of , and Vermamoeba, permit L. pneumophila replication and have been isolated Impact Statement from Legionella-containing waters [21–25]. Based on the In Gram-negative bacteria, the type II secretion system is results of assays done in the laboratory, L. pneumophila notable for its wide-reaching impact on bacterial physi- replicates and/or survives within at least 11 other genera of ology, ecology and pathogenesis. This is especially true protozoa, including Balamuthia, Ciliophrya, Dictyostelium, Echinamoeba, Hartmannella, Oxytricha, Paramecium, Stylo- for Legionella pneumophila, the agent of Legionnaires’ nychia, , Tetramitus (formerly Vahlkampfia) and disease. While giving an update on all aspects of type II [14, 20]. During human infection, L. pneumophila secretion, this review provides a genomic assessment of primarily grows within resident alveolar in the the secretion system across the genus Legionella as well infected lung [26]; however, intracellular infection of type I as hypotheses on how its evolution has been driven by and II alveolar epithelial cells may also contribute to the patho- bacterial interactions with amoebal host cells and other genesis of Legionnaires’ disease [27, 28]. Within , environmental microbes. whether amoebae or macrophages, L. pneumophila evades fusion with and instead modulates endoplasmic reticulum (ER)-to-Golgi vesicular trafficking to remodel the nascent into an ER-derived compartment known containing a signal sequence are first translocated across as the Legionella-containing vacuole [29]. For its intracellular the bacterial inner membrane via the Sec pathway (Fig. 1a) lifestyle, L. pneumophila employs a type IV secretion system [48, 49]. In the periplasm, the proteins are folded into their (T4SS), the Dot/Icm type IVB system, to deliver >300 proteins tertiary conformation, and are destined for translocation (effectors) into the cytosol of infected cells and directly target across the outer membrane via a multiprotein apparatus, host processes including autophagy, death pathways, protein the T2SS [50, 51]. In some instances, nascent proteins that translation and turnover, as well as innate immunity [30]. fold within the and are moved across the inner L. pneumophila encodes a second T4SS, the Lvh type IVA membrane via the twin-arginine translocon (Tat) can be system, which is similar to the Vir T4SS of Agrobacterium recognized by and secreted via the T2SS apparatus [49]. In tumefaciens identified [31]. Although the VirD4 coupling L. pneumophila and a variety of other Gram-negative bacteria protein within the Lvh apparatus has been implicated in [43, 52], the T2SS is composed of 12 ‘core’ components that bacterial entry into host cells and the subsequent evasion are required for biogenesis of the apparatus and secretion of of phagosome acidification, no secreted effectors have yet substrates (Fig. 1a). Four inner membrane proteins (T2S C, been identified [32]. L. pneumophila also has a functional T2S F, T2S L, T2S M) form an assembly platform (AP) to type I secretion system; however, this system is not required which a cytoplasmic ATPase (T2S E) binds [52–57]. After for intracellular growth, although it does enhance bacterial being processed by an inner membrane peptidase (T2S O), a entry into host cells via its secretion of an RtxA-like toxin major pseudopilin (T2S G) and four minor pseudopilins (T2S [33]. The bacterium also secretes a siderophore (rhizoferrin) H, T2S I, T2S J, T2S K) assemble into an envelope-spanning and a melanin-like pigment, both of which promote pilus-like structure [58–61]. The T2S G protein interacts acquisition and, in the case of rhizoferrin L. pneumophila with T2S L, and this interaction is thought to promote pseu- growth in the lungs [34–37]. However, another major facet dopilus biogenesis [62]. Powered by the T2S E ATPase, the of the natural history and pathogenesis of L. pneumophila is pseudopilus appears to act as piston or an Archimedes screw the Lsp type II secretion system (T2SS) [38–42]. Combining to push folded substrates through a homomultimeric secretin experimental data obtained from studies done on L. pneu- pore (T2S D) in the outer membrane and thereby complete mophila with the recent explosion in the genomic database, this review provides an up-to-date assessment of the impact the secretion of the substrates into the extracellular milieu of T2SSs across the genus Legionella, with added attention [48, 50, 52, 63] (Fig. 1a). The T2S C protein links the AP given to the variations in output of the Lsp system as well as and outer membrane components [44, 64, 65] and appears to thoughts on the evolution of this important secretion system. have a substantial role in substrate recognition [44, 66–70]. Since the L. pneumophila system represents one of the most Compatible with secretion occurring in a species-specific well-characterized T2SSs [43–45], the topics and concepts manner, T2S C is among the least conserved proteins amongst covered in this review may be helpful for the evaluation of the various T2SSs; for example, in Vibrio and Dickeya species, T2SS in other bacterial genera. the protein possesses a PDZ domain, whereas in Pseudomonas spcies, it has a coiled-coil domain, and in L. pneumophila, a shorter T2S C has no known domain at its C terminus General overview of the bacterial [44, 67, 69, 71, 72]. Biochemical and structural studies further T2SS suggest that T2SS effectors may also directly interact with T2S L and T2S M in the AP, as well as with the minor pseudopilins Mechanism of protein secretion by the T2SS and the secretin T2S D [70]. The signal contained within the First described in [46, 47], type II secre- substrates themselves that is recognized by the T2SS remains tion (T2S) is a two-step process for secreting proteins into the poorly defined, although proteins secreted by T2S are often extracellular space. During T2S, unfolded protein substrates rich in β-strands [64].

2 White and Cianciotto, Microbial Genomics 2019;5

Fig. 1. Overview of L. pneumophila T2SS. (a) Proteins containing a secretion signal peptide are first translocated across the inner membrane (IM) by the general secretory pathway (Sec) or the twin-arginine translocation pathway (Tat) (not shown). In the periplasm, the signal peptide is cleaved off, and the protein is folded into its tertiary form, and finally secreted into the extracellular milieu by the T2SS apparatus. The T2SS apparatus consists of inner transmembrane proteins (T2S F, L, M), which provide a platform for T2S E to bind. T2S E is a cytoplasmic ATPase which generates energy required to push proteins through the outer membrane (OM) secretin pore (T2S D). T2S O processes the major (T2S G) and minor (T2S H, I, J, K) pseudopilins before they are integrated into the T2SS apparatus, forming a pilus-like structure. T2S C links the inner and outer membrane components and facilitates substrate recognition in the periplasm. (b) Schematic of the five genomic loci encoding Lsp proteins. The distinct loci are separated by double slashes. The individual T2SS genes are indicated by the unique letter associated with the corresponding protein (e.g. D refers to the gene encoding the LspD/T2S D protein) and are coloured to match the colour of the corresponding protein in (a). Promoters are indicated by the black L-shaped arrows. Non-coding RNAs (ncRNA) are indicated by small hatched arrows with the lppnc designation corresponding to the ncRNA found in L. pneumophila strain Paris [73]. Linked genes that do not encode components of the T2SS appear in light grey. All gene arrows are drawn to scale. (c) Overview of gene names and ORF designations for the various T2SS components of L. pneumophila. 130b: L. pneumophila strain 130b; Phil-1: L. pneumophila strain Philadelphia-1.

3 White and Cianciotto, Microbial Genomics 2019;5

In some Gram-negative bacteria, a lipidated protein called of various spanning 15 genera suggests PulS/OutS or the ‘pilotin’ is required for the proper transport the T2SS is absent from this class of (Table S1, and targeting of the T2S D secretin to the outer membrane available in the online version of this article). Some phyla may [74]. However, a canonical pilotin has not been found to be possess T2SSs that deviate from the canonical T2SS found encoded within the L. pneumophila genome, based upon the among Proteobacteria; for example, secretion has been linked use of a hidden Markov model (HMM) [75] to search for to T2SS-like genes in Chlamydia trachomatis and Cytophaga homologues of PFAM ID PF09691 [44, 76]. An alternative hutchisonii yet their genomes lack a complete set of T2SS genes pilotin, AspS, has been described for Vibrio-type T2SSs [77], [43, 97, 98]. Some bacteria, including strains of Escherichia although an HMM search using PFAM ID PF16549 also coli, , Stenotrophomonas maltophilia failed to return any significant hits within theL. pneumophila and , possess two or more distinct T2SSs genome. The apparent absence of a pilotin could suggest [43, 99]. The number of proteins secreted via T2S varies that the Legionella T2S D secretin (also known as LspD, see among the different bacteria, ranging from one inK. oxytoca below) is capable of directing itself to the outer membrane, to >60 in L. pneumophila and Acinetobacter nosocomialis [43] as has been proposed for the secretins of Pseudomonas and (see below). The substrates of T2S are generally delivered into Xanthomonas T2SSs [78]. While the Pseudomonas aeruginosa the extracellular milieu; however, in a minority of cases, they HxcQ and Xanthomonas campestris XpsD possess both a Type can associate with the bacterial cell surface [45]. T2SSs often II/lipoprotein signal peptide and an N-terminal lipobox motif tend to secrete degradative enzymes such as and [78], the DOLOP server [79] suggests that the Legionella peptidases, lipases, and carbohydrate-degrading enzymes that secretin lacks these lipoprotein features; thus, transport of presumably aid in nutrition acquisition, among other things the Legionella secretin to the membrane may be accomplished [43]. In addition to promoting the survival of numerous via a novel mechanism. The Legionella secretin does, however, environmental bacteria, T2SSs can enhance the possess a predicted peptidoglycan-binding SPOR domain at attributes of animal pathogens (e.g. , the N terminus, which may facilitate its localization to the cell , Burkholderia pseudomallei, Chlamydia membrane, as does the peptidoglycan-binding PilQ secretin trachomatis, E. coli, , L. pneumophila, of P. aeruginosa [80, 81]. L. pneumophila also does not possess Photobacterium damselae, P. aeruginosa, S. maltophilia, equivalents of T2S N (sometimes generally referred to as V. cholerae, V. vulnificus and Y. enterocolitica) and plant GspN), T2S A (GspA) or T2S B (GspB), proteins that are vari- pathogens (e.g. species of Dickeya, Erwinia, Pectobacterium, ably present across the T2SSs and are generally dispensable Ralstonia, Xanthomonas and Xylella) [43, 50, 100–109]. for secretion function [82–84]. Overall, the bacterial T2SS is evolutionarily related to type IV pili (T4P) [85], and the T2S O protein is required for both T2S and T4P biogenesis in Role of the L. pneumophila T2SS in the L. pneumophila and others [40]. environment and in disease Genome organization of the Legionella T2SS Many studies have assessed the role of T2S in L. pneu- mophila fitness and growth in both the environment and Initially based on the sequencing of the clinical isolates Phila- within eukaryotic host cells [43]. Most of these studies were delphia-1, Paris, Lens, Corby, Alcoy and 130b (also known conducted by comparing the phenotype of the clinical isolate as strain Wadsworth or AA100) [86–90], the genes encoding strain 130b to that of a mutant specifically lacking a compo- the L. pneumophila T2SS are present within five distinct nent of the T2SS, such as the T2S D secretin (LspD), T2S E chromosomal loci (Fig. 1b, c). This is in contrast to the T2SS ATPase (LspE) or the T2S F inner membrane platform protein of most other organisms that possess T2SS genes encoded (LspF) [42]. Importantly, all 130b mutant phenotypes were within a single operon [91–95]. Promoter analysis and reversed when an intact copy of the T2S protein gene was transcriptional-start-site mapping in L. pneumophila strain re-introduced into the mutant, thereby confirming the role of Paris [73] confirmed that lspF is monocistronic, whereas the the T2SS. While the T2SS mutants grow and survive similarly other lsp genes are co-transcribed with other genes (Fig. 1b). to wild type when inoculated onto solid media [e.g. buffered- ThelspC gene is the first gene in a two-gene operon, with the yeast-extract (BYE) agar] or into liquid bacteriological media second gene encoding a Sel-1 repeat-containing protein that (e.g. BYE broth) at 30 and 37 °C, they are substantially possesses a secretion signal peptide. The lspO (pilD) gene is impaired for growth in the amoebal hosts Acanthamoeba co-transcribed with the T4P-associated genes pilB and pilC, as castellanii, Vermamoeba vermiformis, Naegleria lovaniensis in Aeromonas hydrophila and others [96]. Strand-specific total and Willaertia magna at 35–37 °C [25, 38–40, 42, 112, 113]. RNA sequencing of strain Paris also revealed cis-encoded, The importance of T2S for infection of acanthamoebae has anti-sense RNAs within the lspFGHIJK gene cluster [73]. also been documented through the analysis of an lsp mutant of strain Philadelphia-1 [38]. The T2S mutant growth defect T2SS in other Gram-negative bacteria in the amoebal hosts is several orders of magnitude, and the Many T2SSs have now been characterized, mostly within numbers of mutant bacteria only increase ~1 log after 72 h numerous (but not all) genera in the Alpha-, Beta-, Gamma- of co-culture compared to the numbers of wild-type bacteria and Deltaproteobacteria, although they have also been detected that increase 3–4 log over the same period. This intracel- outside of the Proteobacteria [43, 50] (Fig. 2). Genome analysis lular growth defect becomes even more pronounced when

4 White and Cianciotto, Microbial Genomics 2019;5

Fig. 2. Distribution of T2SS genes among the Proteobacteria and beyond. An unrooted maximum-likelihood phylogenetic tree of Proteobacteria and other bacteria encoding a complete or near-complete T2SS was constructed using aligned 16S rRNA gene sequences [110] in RaxML (100 bootstrap replicates, GTR+Γ model) [111]. Bootstrap support values >50 are presented directly on the branches as grey circles, with larger circles corresponding to higher support values. Bar, 0.1 nucleotide substitutions per site. Clades are colour- coded by class of Proteobacteria, identified by the respective Greek symbols. Bacterial genera that encode a functional T2SS are coloured in green. Genera that are predicted to encode a T2SS but without demonstrated functionality are coloured in black.

the bacterial–amoebal co-cultures are incubated at 22–25 [118]. L. pneumophila T2S also promotes, albeit indirectly, °C [114]. L. pneumophila T2S mutants display difficulty the production of surfactant and thereby facilitates sliding either growing on an agar medium at 25 °C and below or motility on semi-solid agar [119–121]. It is quite likely that surviving planktonically in tap water at low temperature this function of T2S further facilitates the spread and survival [114, 115]. Thus, T2S promotes the environmental persis- of L. pneumophila within the environment. tence of L. pneumophila within intracellular niches and in the planktonic phase over a range of ambient temperatures. Turning our attention to aspects of disease, L. pneumophila T2S may also support environmental persistence by contrib- mutants lacking the T2SS display reduced replication (~10- uting to long-lasting colonization of biofilms; for example, a fold) during intracellular infection of human macrophages, mutant lacking the T2S-dependent substrate Lcl (see below) including the U937 and THP-1 cell lines and mononuclear is impaired in biofilm formation on glass or polystyrene cells obtained from human volunteers [42, 122, 123]. Similar surfaces within static cultures [116, 117]. In support of this, results were obtained when the infection assays used a murine a mutant lacking the T2S O pre-pilin peptidase is unable to alveolar macrophage cell line (MH-S) or bone-marrow- persist within biofilms formed in a dynamic flow-cell system derived macrophages obtained from A/J mice [124]. L. pneu- mophila mutants lacking T2S also exhibit reduced growth

5 White and Cianciotto, Microbial Genomics 2019;5

within human A549 type II epithelial cells and WI-26 VA4 T2SS substrates are present within outer membrane vesicles type I epithelial cells as well as in the murine alveolar epithe- (OMVs) (Table 1). As has been reported for other bacterial lial cell line TC-1 [122, 124]. Within both human and murine T2SSs, such a locale is a result of the substrates existing within macrophages, T2S, although not needed for entry or evasion the periplasm prior to completion of the secretion process as of the , is required for optimal Rab1B association well as occurring, in some cases, on the bacterial cell surface with the Legionella-containing vacuole (LCV) and subsequent [137–139]. An expanded proteomic analysis of supernatants intravacuolar growth between 8 and 12 h post-infection [125]. obtained from cultures of L. pneumophila strain Philadel- Relatively early in the intracellular infection cycle, at least phia-1 and its derivative strain JR32 [132–134] has determined some of the T2S substrates translocate out of the LCV and that another 47 putative substrates containing signal peptides reside nearby in the macrophage cytoplasm [126]. In the A/J are in fact secreted proteins (Table S2). Thus, the number of mouse model of pneumonia, L. pneumophila T2S mutants are T2SS substrates produced by L. pneumophila is likely to be severely impaired in the ability to cause disease and showed at least 72. no evidence of replication within the lungs [42]. Since the Although additional work is needed to confirm the T2SS- T2S mutant is only modestly impaired during in vitro infec- dependency of the 47 candidate effectors, the studies that tion of macrophages and epithelial cells, this observation were mainly done using strain 130b have characterized 25 suggests that the L. pneumophila T2SS promotes processes bona fide substrates of L. pneumophila T2S (Table 1). The in addition to intracellular infection [124]. Compatible with location of the genes encoding these known T2SS substrates such a scenario, the T2SS is also required for a dampening of is shown in Fig. 3. It is apparent that the T2SS effector genes the innate immune response of macrophages that is induced are scattered around the strain 130b as opposed via the MyD88 and Toll-Like Receptor 2 signalling pathways to being localized to one or a few loci or a genomic island. A [123, 124]. In support of these data concerning the various similar conclusion can be made from analysing the genomes defects exhibited by T2SS mutants as compared to the parental of strains Philadelphila-1, Paris and Lens. strain, qRT-PCR analysis and other gene expression and tran- scriptome analyses have confirmed that the T2SS apparatus genes are significantly expressed by wild-type L. pneumophila Degradative enzymes and enzyme activities during growth in bacteriological media and upon intracel- Early studies of L. pneumophila revealed an abundance of lular infection of both macrophages and multiple types of extracellular enzyme activities, including chymotrypsin-like amoebae [25, 76, 113, 115, 125, 127–130]. activity [165], caseinase and gelatinase [166, 167], serum protein degrading [168], aminopeptidases [169], phosphatase, lipase, deoxyribonuclease, ribonuclease, cellu- Secreted substrates (effectors) of lase as well as starch hydrolysis [170]. It was later appreciated the L. pneumophila T2SS that many of these activities are secreted via the T2SS in L. pneumophila strain 130b, based on the reductions in activity Bioinformatic analysis of the genome of strain Philadelphia-1 that were observed in both lsp mutant and effector mutant revealed at least 60 putative substrates of the L. pneumophila supernatants [38–40, 76, 143, 171, 172] (Table 1). More T2SS, i.e. proteins that contain a signal sequence and are limited mutant analysis done with strains Corby and JR32 predicted to have an extracellular localization [131]. Based confirmed the T2S-dependency of some of these enzymes on both a proteomic comparison of culture supernatants across L. pneumophila strains [38, 142, 173]. obtained from wild-type L. pneumophila strain 130b versus an lspF mutant and assessments of enzyme activities in When mutants lacking individual exoenzymes were wild-type versus mutant supernatants, 25 secreted proteins/ analysed in infection assays, the ProA protease, SrnA ribo- activities of strain 130b were confirmed as being dependent nuclease, PlaC acyltransferase and LapA aminopeptidase upon T2SS (Table 1). In most cases, the proteins were later proved to be required for optimal infection of amoebae, also detected in culture supernatants of other strains of and interestingly the relative importance of each of these L. pneumophila [132–134]. The vast majority of these effectors varied depending upon the type of confirmed T2SS substrates contain a typical signal sequence, infected (Table 1). It is surmised that the degradation of indicating that they are moved across the inner membrane by amoebal proteins, peptides, RNA and lipids by these T2SS the Sec translocon prior to incorporation into the T2SS [122]. effectors promote nutrient (e.g. amino acids, nucleotides, The only exceptions are the phospholipase C PlcA and the phosphate, fatty acids) acquisition for intracellular growth, putative peptidyl-proline cis/trans-isomerase (PPIase) LirB, although other scenarios, such as enzyme-mediated modi- which contain a twin-arginine motif and a twin-lysine motif, fications to the LCV, are also possible [76, 158]. ProA is respectively, in their signal peptides and are translocated also notable for being required for the cleavage and acti- across the inner membrane via Tat rather than Sec [134, 135]. vation of the T2SS effectors LapA, LapB, PlaA and PlaC Interestingly, the secretion (or activation) of another phos- [76, 174, 175]. Indeed, the defect that the proA mutant pholipase C activity, which is probably due to the PlcA- exhibits in V. vermiformis is linked to the role ProA has in related PlcB [124], is dependent upon a surface-associated PlaC activation [25]. However, the proA mutant’s defect in PPIase known as Mip [136]. In addition to being detected N. lovaniensis is independent of the protease’s activation ‘free’ within culture supernatants, a number of the validated of LapA, LapB, PlaA or PlaC, suggesting that ProA may

6 White and Cianciotto, Microbial Genomics 2019;5 Continued 149] 144] 131, 132] 131, 147] 131, 147] 145, 146] References [25, 76, 113, [25, 113, 124, [76, 113, 124, [25, 113, 124, [25, 113, 124, [25, 113, 124, [25, 113, 142] 124, 131, 143, 131, 132, 141] 131–133, 143] [116, 131, 132, [133, 134, 148, [131, 133, 140] [113, 131, 132]

] ] ] ] ] ] ] ) ) ) (42 % (40 % ] −92 −74 −74 ) ) ] (36 % I, ) ) ) ) ) ) (40 % I, (55 % I, (43 % I, −89 −66 −70 −10 −89 −40 −145 −109 (56 % I, spp. (24 % I, spp. Flagellimonas ] Flagellimonas [Fungi] E=1×10 E=5×10 E=1×10 E=2×10 E=1×10 E=9×10 homologue [ Nitrospirae I, E=2×10 I, E=8×10 [ Lentisphaerae (67 % I, E=2×10 (66 % I, E=3×10 (49 % I, E=3×10 aquimarina [ [ Alphaproteobacteria [ [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria ClosestLegionella non- Aquicella lusitana Aquicella lusitana Aquicella lusitana Leptospirillum ferriphilum Sphingorhabdus flavimaris Victivallis vadensis Victivallis Methylococcus Berkiella cookevillensisBerkiella [ Bacteroidetes Poseidonocella sedimentorum Poseidonocella Spizellomyces punctatus 89 % 16 % 84 % 39 % 53 % 95 % 74 % 11 % 67 % 54 % 47 % within Legionella Prevalence Crystal structure PDB: 6ESL PDB: PDB: 5GNE PDB: Role in infection ‡ in infection Role not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine may promote growth in A549, Ac, in A549, Ac, growth promote may Vv U937 and not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine promotes growth in murine lung in murine growth promotes in A549, growth for required not Wm and Nl, U937, Vv Ac, promotes growth in Ac growth promotes in A549, growth for required not murine and Wm Nl, U937, Vv, lung not required for growth in Ac, Nl, in Ac, growth for required not Wm and U937, Vv promotes attachment to A549, to attachment promotes U937 and NCI-H292 and attachment promotes Vv and Ac of invasion in Ac growth for required not not required for growth in Ac in Ac growth for required not HL-60 and not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine not required for growth in Ac, Nl, in Ac, growth for required not Wm and U937, Vv • • • • • • • • • • • • • • • Sup’t Sup’t Sup’t Sup’t Surface substrate) Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, Sup’t, OMV, OMV, Sup’t, Location(s) † Sup’t, OMV (Tat (Tat OMV Sup’t, -isomerase protein novelty chitinase endoglucanase trans aminopeptidase putative amidase putative triacylglycerol lipase eukaryotic-like lys/arg eukaryotic-like lys/arg monoacylglycerol lipase monoacylglycerol ile/met/asp aminopeptidase ile/met/asp eukaryotic-like collagen-like eukaryotic-like collagen-like putative peptidyl proline cis- peptidyl proline putative Protein activity or sequence activity or Protein eukaryotic-like glucoamylase eukaryotic-like leu/tyr/phe/val/ eukaryotic-like putative protease eukaryotic-like putative 1 ORF lpg0264 lpg1918 lpg1116 lpg2814 lpg0468 lpg0422 lpg0032 lpg2644 lpg1962 lpg1157 lpg2999 Strain Phil- Strain ORF lpw03521 lpw19571 lpw11641 lpw30701 lpw05481 lpw05041 lpw00321 lpw28961 lpw20131 lpw12111 lpw32851 Strain 130b Strain AmiA T2SS substrate CelA ChiA LapA LipA GamA LapB Lcl LirB LipB LegP * pneumophila of L. T2SS substrates Documented 1. Table

7 White and Cianciotto, Microbial Genomics 2019;5 Continued 152] 147] 151] [76, 131] [131, 154] [131, 153] [113, 131] References [25, 76, 113, [25, 113, 124, [25, 113, 124, [25, 113, 124, [25, 113, 131, [25, 113, 131] 131, 132, 134, 131, 132, 150, 131, 155, 156] 124, 132, 157] [131, 132, 153]

] ] ] ] ] ] ] ) (32 % I, −33 ) (34 % I, ) ) ) ) ) ) (44 % I, (25 % I, −48 −81 −22 −11 −19 −47 −131 spp. (39 % I, spp. None None None E=6×10 E=3×10 E=1×10 E=5×10 E=2×10 E=3×10 homologue I, E=2×10 [Cyanobacteria] (27 % I, E=7×10 Francisella Dyella japonica [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria ClosestLegionella non- Aquicella lusitana Nostoc punctiforme (35 % salmonis Piscirickettsia (37 % I, litoralis Piscirickettsia Francisella halioticida Francisella Parendozoicomonas haliclonae Parendozoicomonas 58 % 49 % 65 % 91 % 77 % 86 % 84 % 16 % 77 % 75 % 100 % within Legionella Prevalence Crystal structure PDB: 5XTA PDB: PDB: 6A0N PDB: PDB: 4KH9 PDB: PDB: 5CDH PDB: Role in infection ‡ in infection Role not determined not not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine promotes destabilization of the of destabilization promotes LCV in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine may promote growth in Ac, Nl, Nl, in Ac, growth promote may Vv U937 and in Ac growth promote may Vv Nl, in Ac, growth promotes and Wm in A549 growth for required not U937 and promotes growth in Vv and Wm and in Vv growth promotes in Ac growth promote may in Nl growth for required not promotes growth in Ac growth promotes in Nl, growth for required not Vv U937 and not required for growth in A549, growth for required not and Wm Nl, U937, Vv, Ac, lung murine promotes growth in Ac and Wm and in Ac growth promotes in Nl, growth for required not Vv U937 and not required for growth in Ac, Nl, in Ac, growth for required not Wm and U937, Vv • • • • • • • • • • • • • • • • • Sup’t Sup’t Sup’t Sup’t Sup’t Sup’t Sup’t substrate) Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, OMV Sup’t, Location(s) † Sup’t, OMV (Tat (Tat OMV Sup’t, novel novel novel novel protein novelty novel, VirK-like novel, phospholipase A acid phosphatase lysophospholipase A glycerophospholipid: glycerophospholipid: Novel C1 family peptidase C1 family Novel novel, DUF4785-containing DUF4785-containing novel, Protein activity or sequence activity or Protein cholesterol transferase (GCAT), (GCAT), transferase cholesterol eukaryotic-like phospholipase C eukaryotic-like phospholipase eukaryotic-like tartrate-sensitive eukaryotic-like tartrate-sensitive 1 ORF lpg1119 lpg1832 lpg2343 lpg0189 lpg0873 lpg2837 lpg1809 lpg0956 lpg0502 lpg1385 lpg2622 Strain Phil- Strain ORF lpw11671 lpw18641 lpw25361 lpw02811 lpw09571 lpw30971 lpw18401 lpw10421 lpw05821 lpw13951 lpw28721 Strain 130b Strain Continued

Map T2SS substrate NttG PlaA NttE NttF PlaC NttC NttD PlcA NttA NttB Table 1. Table

8 White and Cianciotto, Microbial Genomics 2019;5 132] 131–133, 131, 158] 158–164] References [25, 113, 124, [25, 113, 124, [25, 113, 124, ] ] ] ) (40 % −63 ) ) (43 % I, −82 −138 E=2×10 homologue I, E=1×10 (38 % I, E=2×10 [ Gammaproteobacteria [ Gammaproteobacteria [ Gammaproteobacteria ClosestLegionella non- fluorescens Pseudomonas Aquicella lusitana Francisella philomiragia Dictyostelium discoideum Dictyostelium Willaertia magna ; Dd, 95 % 33 % 100 % within Legionella Prevalence Crystal structure ; Wm, vermiformis ; Wm, Vermamoeba Role in infection ‡ in infection Role promotes growth in Nl and Vv in Nl and growth promotes in A549, growth for required not lung murine and U937, Wm Ac, not required for growth in A549, growth for required not Wm and Nl, U937, Vv Ac, promotes tissue destruction tissue in promotes lung Vv in Nl and growth promotes in Ac growth promote may in A549, growth for required not explanted U937, Wm, HL-60, and macrophages, guinea pig lung murine • • • • • • • Sup’t Sup’t, OMV Sup’t, Sup’t, OMV Sup’t, Location(s) † novelty metalloprotease T2 ribonuclease Protein activity or sequence activity or Protein eukaryotic-like phospholipase C eukaryotic-like phospholipase 1 ORF lpg2848 lpg1455 lpg0467 Strain Phil- Strain ORF lpw31111 lpw14741 lpw05471 Strain 130b Strain Continued

SrnA PlcB T2SS substrate ProA *Based on the presence of the indicated protein in wild-type culture supernatants and absence in T2SS mutant culture supernatants, plus the presence of a secretion signal at the N terminus of a secretion plus the presence supernatants, T2SS mutant culture supernatants and absence in in wild-type culture protein of the indicated *Based on the presence protein. of the predicted Tat, to be predicted that are Proteins surface. on the bacterial cell also present Surface, vesicles; in outer membrane also present supernatant; OMV, culture in broth is present protein †Sup’t, in parentheses. indicated are substrates than Sec, rather when the mutant wild-type; ‘promotes’, from not different when the mutant was ‘not required’, assay(s): infection mutant(s) in the indicated of the corresponding ‡Based on the behaviour been but genetic complementation has not yet impaired when the mutant was promote’, genetic complementation; ‘may by reversed was and that defect to wild-type, relative impaired was ; Naegleria lovaniensis polyphaga ; Nl, Acanthamoeba Vv,; Ap, castellanii Acanthamoeba Ac, attempted or achieved. Table 1. Table

9 White and Cianciotto, Microbial Genomics 2019;5

Fig. 3. Chromosomal organization of T2SS genes in L. pneumophila strain 130b. The entire chromosome is depicted as a circular map. The tick marks indicate the nucleotide position from 0 to ~3 500 000 bp along the circular chromosome, with 0 indicating the origin of replication. From outside in, the two bands (in aqua) depict the predicted coding sequences transcribed clockwise and anticlockwise, respectively. The next band inward indicates the genomic positions of known individual T2S effectors (blue lines, with gene names nearby), putative T2S effectors (grey) and T2SS apparatus genes (red lines, with gene names nearby). Further inside is the GC content, with gold indicating above average and purple indicating below average GC content relative to the genomic average (38.2 mol%). The inner-most band represents the GC skew, indicative of the preference for G (grey) or C (orange) base pairs.

activate additional T2SS effectors or in some cases directly pneumophila growth in macrophages, suggesting functional target the host to promote bacterial replication. In the redundancy among the effectors and/or the existence of case of LapA, the crystal structure of the T2SS effector has other T2SS substrates that are more critical for infection been recently determined, providing insight into the broad of mammalian cells [124, 176–178]. A limited analysis has specificity of this aminopeptidase, which is active against also failed to uncover a T2SS substrate that is required for >10 substrates [76] (Table 1). Incidentally, other known intracellular growth in epithelial cells [124]. Intriguingly, the T2S-dependent exoenzymes that have had their structures ChiA chitinase is needed for full bacterial growth in the lungs resolved include the LapB aminopeptidase [76, 144] and of infected A/J mice [131]. Since a chiA mutant appears to Map acid phosphatase [150, 151] (Table 1). be normal for growth in macrophages and epithelial cells, Thus far, no known T2SS substrate, whether an exoen- it is not clear how the chitinase promotes intrapulmonary zyme or not, has been documented as being required in L. growth. However, given that mammals do not possess chitin,

10 White and Cianciotto, Microbial Genomics 2019;5

the T2SS effector must be degrading a ‘chitin-like’ molecule in seemingly restricted to the genus Legionella as in the case of the lungs and/or encoding another type of activity. Although NttA, NttC and NttE (Table 1). ChiA is the only T2SS effector that has been shown to be required for bacterial survival in the lungs, ProA probably Eukaryotic-like domains within T2SS substrates also contributes to disease by mediating the destruction of lung tissue [159, 161–164, 179]. Additionally, ProA may aid Since the completion of the L. pneumophila genome, in both iron assimilation by degrading host transferrin and eukaryotic-like domains have been a hallmark of the effec- immune evasion by degrading cytokines [124, 180]. tors of the Dot/Icm T4SS, and it has been hypothesized that eukaryotic-like T4SS effectors were acquired via inter- For three reasons, we suggest that the L. pneumophila T2SS kingdom horizontal gene transfer (HGT) [86, 187–190]. It elaborates multiple other secreted enzymes. First, while many is important to emphasize that eukaryotic-like domains also secreted activities are completely abolished upon muta- often exist among the known T2SS effectors. As reported in tion of the corresponding substrate gene, there is residual 2001, the first such-characterized substrate is the histidine chitinase activity in a chiA mutant and residual aminopepti- acid phosphatase Map (150). Although Map is most closely dase activity in a lapA lapB double mutant, suggesting the related to a histidine-type phosphatase of Francisella spp. existence of additional secreted enzymes with overlapping (Table 1), phylogenetic analysis reveals that the Legionella functions [131, 143]. Compatible with these observations, (and Francisella) protein is most closely related to eukary- L. pneumophila encodes two more putative aminopeptidases otic acid phosphatases such as those found among genera of [Lpw05621 (Lpg0482) and Lpw12101 (Lpg1156)] and one red algae, including Gracilariopsis, Chondrus and Galdieria additional putative chitinase [Lpw24031 (Lpg2217)] (Table (Fig. 4a). Interestingly, Francisella species, the only other S2). Second, there are enzyme activities present in wild-type, bacteria possessing a protein that is highly similar to Map, but not lsp mutant, supernatants that have not yet been linked are, like Legionella species, capable of replication within to a known T2S substrate. These activities include tartrate- macrophages and protozoa [191–194]. Thus, the HGT of Map resistant acid phosphatase, diacylglycerol lipase, peptidoglycan from a eukaryotic host(s) may have facilitated the acquisition hydrolase, xylanase and DNase [140, 147, 150, 170, 171]. of some aspect of the intracellular lifecycle of these bacteria, In line with these results, L. pneumophila secretes another although no infection phenotype has been described thus far putative lipase [Lpw10431 (Lpg0957)] as well as a putative for a map mutant [25, 113, 150]. xylanase [Lpw07891 (Lpg0712)] (Table S2). The lsp mutants of L. pneumophila strain 130b are also impaired for surfactant Originally identified as a eukaryotic-like protein based on the production, sliding motility and poly-3-hydroxybutyrate presence of an amylase domain [86, 187], GamA is involved metabolism, suggesting the existence of yet additional T2SS in the breakdown of the eukaryotic storage molecule glycogen effectors that may have enzymatic activity [120]. Third, based [142]. Although GamA-like proteins with signal sequences upon proteomic analysis of strain 130b, there are documented are found among Gamma-and Deltaproteobacteria, blastp T2SS substrates that have strong sequence similarity to analysis reveals that GamA is most closely related to a protein known enzymes in other bacteria (Table 1). These include of the fungus Spizellomyces punctatus (Table 1). Phylogenetic both the LirB protein, which is a putative PPIase that is highly analysis of the most closely related GamA homologues span- expressed at low temperatures [149], and the AmiA protein, ning 20 genera confirms that GamA is most closely related to which is likely to be an amidase [131]. For AmiA, Phyre2 eukaryotic proteins (Fig. 4b). Despite the relatedness of GamA analysis [181] identified, with 100 % confidence, the AmpD to eukaryotic proteins, gamA mutants are not impaired for amidase from [182] as the top template intracellular infection [25, 142], suggesting that the protein to model the tertiary structure of ~82 % of the AmiA protein has a dispensable role in L. pneumophila growth within host (31 % identity over 168 residues, from amino acids 33 to 200). cells. In a similar vein, functional annotation of the secretome of blastp analysis reveals that PlcA is most closely related to a strain Philadelphia-1 suggests that 20 % of the T2SS effectors putative phospholipase in Aquicella lusitana, whereas PlcB is are peptidases [183], which is compatible with L. pneumophila most akin to a putative phospholipase in Pseudomonas fluore- having a tendency to use amino acids as its primary food scens (Table 1). PlcA- and PlcB-like proteins are also present [184–186]. in other intra-amoebal parasites (or endosymbionts) such For 15 of the T2SS-dependent exoenzymes, including as other Aquicella and Pseudomonas species as well as Burk- ChiA, PlaC, ProA and SrnA, the protein has its greatest holderia and Caedimonas species [195–197], . Interestingly, homology to proteins/enzymes encoded by various genera however, the phospholipase C proteins PlcA and PlcB, which of Gammaproteobacteria (Table 1), which is not unexpected are 36 % identical and 56 % similar to each other, belong given the position of Legionella within the Gammaproteobac- within the phosphatidylcholine-hydrolysing group of eukary- teria (Fig. 2). Interestingly, however, PlaA is most similar to otic phospholipases C that spans from the yeast Saitoella to proteins encoded within the cyanobacteria, and AmiA, Lcl, the marine corals Acropora, Orbicella and Stylophora [197]. LegP, LipB and LirB, have close bacterial homologues in the Indeed, phylogenetic analysis supports the view that both Alphaproteobacteria or elsewhere (Table 1). Arguably, most PlcA and PlcB are eukaryotic-like (Fig. 4c), although neither interestingly, some T2SS substrates are either most closely protein has been found thus far to be required for intracellular related to a eukaryotic enzyme(s), as in the case of GamA, or infection (Table 1). Interestingly, a triple mutant lacking PlcA,

11 White and Cianciotto, Microbial Genomics 2019;5

Fig. 4. Phylogenetic analysis of select eukaryotic-like T2SS effectors ofL. pneumophila. Homologues of Legionella T2SS effectors were identified byblastp using a minimum query coverage of 60 %, and identity and E-value cutoffs of 30 and 1×10−30, respectively, for panels a–e, and amino acid identity and E-value cutoffs of 25 and 1×10−15, respectively, for panel f. Maximum-likelihood trees were generated from full-length amino acid alignments of the T2SS effectors and the most closely related homologues encompassing at least 20 genera per effector group using RaxML (100 bootstrap replicates, GTR+Γ model) [111]. The trees of related sequences are given for the acid phosphatase Map (a), glucoamylase GamA (b), phospholipases PlcA and PlcB (c), putative astacin protease LegP (d), aminopeptidases LapA and LapB (e), and chitinase domain of ChiA (f). Bootstrap support values >50 are presented at the respective nodes. Bars, 0.1 amino acid substitutions per site. Monophyletic clades of bacterial homologues have been collapsed in panels d (N=70 genera) and f (N=16 genera) for space. Eukaryotes are indicated by red branches and bacteria by blue branches. GenBank accession numbers of the analysed protein sequences are listed before the respective genus designations.

12 White and Cianciotto, Microbial Genomics 2019;5

PlcB and a Dot/Icm T4SS-dependent PLC (PlcC) is impaired chitinase domain from water moulds. Thus, like LapA, ChiA in a Galleria mellonella infection model [173]. may be an example of a T2SS effector that evolved as a result of L. pneumophila growth within amoebae. Another known T2SS effector of L. pneumophila that can be considered eukaryotic-like is LegP (Table 1). LegP contains As just described, 8/25 (32 %) of the known T2SS effectors an astacin-like protease domain [131, 187] and phylogeneti- are eukaryotic-like. Although the known T2SS substrate Lcl cally LegP-like proteins, although found in many bacterial has the Gly-aaX-aaY collagen helix motif found originally Gram-positive and Gram-negative genera, are highly similar in eukaryotes [116, 117, 131, 145], it and other bacterial to putative proteases from marine eukaryotes including collagen-like proteins primarily share similarities with eukar- the cnidarians Nematostella, Pocillopora and Exaiptasia, yotic proteins at the structural level [206]. Furthermore, the as well as the ocean-dwelling worm Saccoglossus (Fig. 4d). amino acids at positions X and Y within the collagen helix of LegP is also unusual for being translocated out of the LCV Lcl are rather distinct from those found in eukaryotes [206]. in a T4SS-dependent manner, while being secreted into the Consequently, we would not consider Lcl to be a ninth eukar- bacterial culture supernatants via the T2SS [131, 198]. Such a yotic-like T2SS effector. However, if one examines the other dual-secretion phenomenon may also apply to several of the 47 putative substrates of the T2SS (Table S2), there are four putative substrates (Table S2). The molecular basis for this additional eukaryotic-like effectors [i.e. Lpw03931 (Lpg0301), differential secretion is unknown. However, it was recently Lpw10571 (Lpg0971), Lpw24081 (Lpg2222) and Lpw28361 documented that can secrete the (Lpg2588)]. This suggests that at least 17 % (i.e. 12/72) of TDH into the extracellular milieu via both T2SS the L. pneumophila T2SS substrates are eukaryotic-like in and a type III secretion system [199], lending support to the nature. Thus, eukaryotic-like effectors of L. pneumophila are existence of dual secretion mechanisms. not restricted to the T4SS. It is posited that bacterial effectors have been acquired by both HGT and convergent evolution LapA and LapB, which are 45 % identical and 63 % similar, [207]. We favour the hypothesis that eukaryotic-like T2SS are aminopeptidases that provide critical nutrients to effectors were acquired via HGT, as HGT is detectable at L. pneumophila during intracellular infection of protozoa the primary sequence level, whereas convergent evolution [76, 143]. LapA and LapB share high sequence homology is more commonly detected at the gross structural level with a secreted aminopeptidase of A. castellanii, which, as [203]. Given that only a few annotated genomes of protozoa noted above, is one of the major hosts for L. pneumophila exist yet amoebae are probably major contributors to HGT, [76]. Phylogenetic analysis affirmed that LapA may have been our understanding of the origins of these eukaryotic-like acquired from a protozoan host, with other amoebal-parasites Legionella proteins is only just beginning. Furthermore, we such as Aquicella, Burkholderia and Duganella species also suspect that the numbers of eukaryotic/protozoan-like T2SS possessing LapA-like proteins (Fig. 4e). On the other hand, substrates will rise substantially as more amoebal genomes LapB represents a more recent gene duplication, with LapB are sequenced. undergoing faster adaptation and possessing enzymatic activities distinct from LapA [76]. That protozoa were prob- Novel effectors ably the direct source of genetic material for Legionella has been previously proposed for many T4SS substrates as well Interestingly, 27 of T2SS effectors encoded by L. pneumophila as some non-secreted proteins [200, 201]. do not share significant structural or sequence similarity to any known enzyme(s). Seven of these novel effectors (i.e. Yet another eukaryotic-like effector is ChiA. Although most NttA, NttB, NttC, NttD, NttE, NttF and NttG) have been closely related to a hypothetical protein in the gammaproteo- validated as T2S substrates, i.e. they are present in wild-type bacterium A. lusitana (Table 1), ChiA possesses a family-18 supernatants but not lsp mutant supernatants (Table 1). The chitinase domain that is most akin to glycosyl hydrolase other 20 (i.e. Lpg0042, Lpg0165, Lpg0198, Lpg0301, Lpg0374, domains encoded by mimiviruses that infect the amoebae Lpg798, Lpg0957, Lpg1030, Lpg1233, LvrE, Lpg1318, A. castellanii and V. vermiformis [202]. Amino acid residues Lpg1431, Lpg1645, Lpg1647, WipC, Lpg2220, Lpg2246, 445–782 of ChiA have high relatedness (56 % identity, 73 % Lpg2275, Lpg2320 and Lpg2443) have been detected in wild- similarity, E-value=3×10−146) to a mimivirus isolated from the type strain Philadelphia-1 supernatants but have not yet been ocean depths, whereas residues 443–782 share high amino- examined for their lack of secretion by the corresponding acid sequence homology (53 % identity, 70 % similarity, lsp mutant (Table S2). In many cases, members of this class E-value 3×10−137) with another mimivirus isolated from a of T2SS substrates share, to varying degrees, sequence simi- high-alkalinity/high-salinity lake. Given that mimiviruses larity to hypothetical proteins in other bacteria. For example, and L. pneumophila have co-evolved with the protozoan host, NttD, possessing the conserved domain of unknown function it is not surprising that mimiviruses have been proposed as (DUF) 4785, has homologues that are found primarily in a source for HGT in Legionella species [189, 203, 204]. From phylogenetically related Gammaproteobacteria. On the other phylogenetic analysis, the chitinase domain of ChiA also hand, NttB has putative homologues only among aquatic appears related to putative chitinases found in water moulds Piscirickettsia species and Silvanigrella aquatica. Recent (Fig. 4f). Since water moulds were previously implicated in structural and biochemical analysis revealed that NttB is a HGT with mimiviruses [205], we posit that mimiviruses may C1 family peptidase that diverged from common papain‐like have been a conduit for L. pneumophila acquisition of the proteases and forms a distinct phylogenetic lineage

13 White and Cianciotto, Microbial Genomics 2019;5

from eukaryotic cathepsins [152]. NttF has only a single studies had found that celA, chiA, legP, map and nttA are homologue found in the genome of Piscirickettsia litoralis. modulated by the regulators PmrA and PmrB [210], whereas Finally, NttG has only a single homologue, and that related lipA and lipB are influenced by LetA and RpoS [129], and protein is encoded by aquatic Francisella halioticida, a gamA is affected by CsrA [211]. The CpxRA two‐component member of a genus that, like Legionella, is pathogenic for both system, which controls expression of the Dot/Icm system and amoebae and humans [208, 209]. Structural analysis suggests effectors, was also shown to positively regulate expression of that NttG is a VirK-like protein, yet its activity and role of 13/25 T2SS effectors [212], including at least six factors (LapA, infection remain undefined [154]. Arguably most interest- NttA, NttD, PlaC, ProA and SrnA) that promote intracellular ingly, some members of this general class of T2S substrates replication in protozoa (Table 1). A comprehensive summary do not have any putative homologues (E-value <1×10−10) of the various regulatory aspects of the T2SS effector genes is outside of the genus Legionella, further suggesting that many presented in Table S3. of the T2SS-dependent proteins may be highly specialized Although proteomics and ensuing mutant analysis has been for the intra-amoebal lifestyle of Legionella species [25, 113]. the principal means by which T2SS-dependent proteins These include the documented T2SS substrates NttA, NttC that promote infection have been identified, transcriptional and NttE (Table 1) as well as the putative substrates Lpg0042, profiling has recently been shown to be a valid alternative. Lpg0374, Lpg0798, Lpg1233 and Lpg2443 (Table S2). Impor- Indeed, the importance of LapA and PlaC for infection of tantly, both of the effectors in this category that have been A. castellanii was revealed through a novel combination of assessed, using mutant analysis, for their role in intracellular transcriptional and mutational analyses [76]; that is, (i) the infection were found to be required for optimal growth two genes were first found to be among the most up-regulated within amoebae. Whereas NttA is necessary for infection of effector genes during wild-type infection of the amoebae, (ii) A. castellanii and W. magna, NttC is required for infection transcript profiling of a lapA mutant then showed even higher of V. vermiformis and W. magna [25, 113]. Given the novelty of levels of plaC mRNA, and conversely a plaC mutant exhibited NttA and NttC, it is difficult predict how these T2S substrates elevated levels of lapA transcription, and (iii) a newly made, promote intracellular infection; however, further phenotypic double mutant lacking both lapA and plaC exhibited a loss analysis of the nttA and nttC mutants as well as biochemical of infectivity, uncovering redundant yet important roles for and structural analysis of the NttA and NttC proteins may LapA and PlaC in nutrient acquisition and intracellular bacte- represent fruitful lines of inquiry. Although not peculiar to rial growth. the genus Legionella because of related hypothetical proteins occurring mainly in Gammaproteobacteria, the novel effector NttD is also required for optimal infection of A. castellanii The T2SS and its effectors belong to [76]. The structure of NttD has been recently obtained; but, the core genome of L. pneumophila unfortunately, this information has not yet provided a strong clue as to the activity of NttD [76]. Given that three of four Although the vast majority of studies on the T2SS have utilized novel effectors examined (i.e. NttA, NttC and NttD; but not serogroup (SG)−1 strain 130b and to a lesser extent SG-1 NttB) promote infection of at least one amoebal host, it is strains Philadelphila-1 and Corby, we and others previously likely that the emergence of novel T2SS substrates plays a reported the presence of T2SS apparatus proteins encoded significant role in both the ecology and the pathogenesis of in a variety of clinical and environmental L. pneumophila L. pneumophila. isolates [42, 76, 213–215]. Extending this analysis to all of the 90 annotated L. pneumophila complete genome assemblies currently in the NCBI Reference Sequence (RefSeq) Database Transcriptional analysis and regulation of T2SS [216], encompassing eight of the 15 SGs, we found that all effectors of the T2SS apparatus genes are intact in all of the strains, Recently, qRT-PCR analysis was used to assess the relative except for frame-shift mutations in lspK in SG1 strain Flint expression of 19 of the 25 known effector genes during multiple 2 (D-7477), lspL in SG1 strain FFI103 and pilD in SG1 strain stages of L. pneumophila growth in bacteriological media as L10/23. The minimum amino-acid identity of the Lsp homo- well as during intracellular replication in three amoebae and logues compared to the L. pneumophila 130b Lsp proteins was human macrophages [76]. The T2SS substrate genes showed 93.5 % for LspC, 96.5 % for LspD, 97.0 % for LspE, 96.5 % a range of expression patterns as opposed to displaying for LspF, 98.6 % for LspG, 90.2 % for LspH, 92.8 % for LspI, similar responses to the various growth environments; for 93.7 % for LspJ, 91.9 % for LspK, 88.7 % for LspL, 94.2 % example, eight of the 19 genes were up-regulated upon intra- for LspM and 89.2 % for PilD/LspO, in agreement with our cellular infection, and eight others were down-regulated [76] previous findings from analysing a panel of 17 strains [76]. (Table S3). Together, these data imply that the amounts of Some apparatus proteins such as LspJ may undergo diver- proteins that are elaborated by the T2SS are dictated, at least sifying selection within L. pneumophila [217], which may primarily, at the level of the individual effector-gene or of help to explain the varying degrees of secreted activity of subsets of effector-gene transcription as opposed to being environmental isolates despite encoding an intact T2SS [218]. controlled at the level of T2SS apparatus gene transcription Turning attention to the prevalence of the secreted substrates, or by a single global regulator that acts upon the many effector it is clear that the genes for all 25 confirmed T2SS effectors genes [76]. In further support of this conclusion, earlier (Table 1) are present and intact within the 90 annotated

14 White and Cianciotto, Microbial Genomics 2019;5

L. pneumophila genomes, with the sole exception being a frame- (lspGHIJK). The second notable difference regarding the shift mutation ingamA in SG1 strain Albuquerque 1 (D-7474). lsp genes among the Legionella genus is the apparent pseu- Overall, these findings suggest that the T2SS along with many dogenization of the lspH gene of L. norrlandica resulting effectors belongs to the core genome of L. pneumophila. While in a truncated LspH protein. Although most closely related the T4SS also belongs to the core genome of L. pneumophila, to L. pneumophila (Fig. 5a), L. norrlandica is avirulent in a it has been reported that up to 30 % of T4S effectors belong protozoan infection model and is unable to establish large to the accessory genome and undergo increased rates of replication [11]. Inactivation of the lspGHIJK pseudogenization [89]. pseudopilin gene cluster in L. pneumophila results in loss of T2S activities in culture supernatants and inability to grow within protozoa [39]. Thus, we hypothesize that the Conservation of the T2SS apparatus attenuation of L. norrlandica, which incidentally possesses genes within Legionella an intact T4SS [11], is attributable to loss of the T2SS via the lspH mutation. Description of T2SS genes in other Legionella species Conservation and context of T2SS genes among As mentioned in the introductory section, 63 species of Legionella species Legionella have thus far been characterized, with 32 of them already being linked to human disease (Fig. 5a). The degree of conservation of Lsp protein sequences is more Moreover, whenever examined, the non-pneumophila variable among the Legionella species than it is amongst species have also proven to be intracellular parasites of L. pneumophila strains. The major pseudopilin LspG displays amoebae [11, 219–224]. Nonetheless, our understanding the highest degree of conservation, with a minimum amino- of these Legionella species, including L. longbeachae, acid identity of 78.8 % among species relative to LspG in which is the most prevalent disease-causing species in L. pneumophila. The minimum amino-acid identity for other , has lagged very far behind that of L. pneumophila Lsp proteins is 77.3 % for LspE, 70.6 % for LspF, 63.1 % for [14, 20, 224, 225]. The presence of T2SS genes in non- LspD, 54.4 % for LspI, 51.5 % for LspJ, 48.2 % for PilD/LspO, L. pneumophila species of Legionella was first detected in 41.8 % for LspH, 40.3 % for LspC, 40.0 % for LspK, 37.8 % L. cherrii, L. feeleii, L. gormanii, L. longbeachae, L. micdadei, for LspM and 34.6 % for LspL. Despite the higher divergence L. parisiensis and L. spiritensis by Southern blot analysis among sequences, only LspD has reportedly undergone diver- [42], prior to the sequencing of any of the non-pneumophila sifying selection within certain clades of the Legionella evolu- species [224, 226]. With the elucidation of many Legionella tionary tree [217], which may play a role in diversification species genomes, we previously confirmed the presence of of T2SS function, and consequently defining the ecological the T2SS apparatus genes among all 41 Legionella species niche of Legionella species. examined [76]. Extending this analysis to include all 57 The chromosomal organization oflsp gene clusters varies of the currently sequenced Legionella genomes, coding across the genus. Currently, at least 12 different arrange- sequences corresponding to all core T2SS genes are present ments of the lsp gene clusters are evident (Fig. 6). Each across the genus (Fig. 5b). While all lsp genes were detected arrangement is most similar among phylogenetically close in all of the 57 species analysed, there were two notable species, such as L. waltersii and L. pneumophila. Although differences in the gene arrangements. First, there were the five lsp gene clusters are found intact across all intergenic insertions between lspF and lspG in the lspF- Legionella species, the genetic context of each cluster varies. GHIJK locus in L. drozanskii, L. maceachernii, L. micdadei In some cases, lsp genes are flanked by conserved ortholo- and L. nautarum (Fig. 5b). All species in this clade had gous genes; for example, T2S O (pilD) was always found as inserted a gene that encodes TesA, a signal sequence- the last gene in the pilBCD operon, and T2S C (lspC) was containing, multi-functional periplasmic protein with always immediately upstream of a Sel1 repeat-containing thioesterase 1/protease 1/lysophospholipase L1 activity protein. In other cases, the flanking genes were highly vari- [227]. Since L. drozanskii, L. maceachernii, L. micdadei able. The genes flanking the lspDE cluster were notable for and L. nautarum are monophyletic (Fig. 5a), the insertion being highly diverse among the analysed Legionella species. event probably occurred once and has persisted since. In L. spiritensis, L. hackeliae and L. clemsonensis, there was Based on blastp analysis of TesA, the source of tesA was a methionine tRNA immediately upstream of T2S D. In likely aquatic bacteria including Polynucleobacter or Vibrio L. oakridgensis, an ISL3 family insertion sequence has trans- spp., and no TesA homolog was found to occur within the posed between the tRNA and T2S D. It is well established L. pneumophila genome. L. maceachernii also had a gene that tRNAs serve as integration sites in various prokaryotes encoding a hypothetical protein inserted between lspF and [229]. In Legionella species, the type IVA secretion system tesA (Fig. 5b). This putative protein lacks a secretion signal is encoded on a plasmid-like element that integrates at the and is found strictly within L. maceachernii based on the 3' end of various tRNAs in both L. pneumophila and L. long- absence of any homologues in the blast protein data- beachae [226]. Thus, the close proximity of the lspDE gene base. As lspG transcription is not linked to lspF (Fig. 1b), cluster to a tRNA gene may explain the high diversity of insertion of genes between lspF and lspG is unlikely to flanking genes observed. Given the various arrangements impact transcription from the pseudopilin gene operon of the five lsp gene clusters on each chromosome, it is clear

15 White and Cianciotto, Microbial Genomics 2019;5

Fig. 5. Phylogenetic analysis and distribution of T2SS genes in Legionellales. (a) A list of all currently named Legionella species, their phylogenetic relationships based upon data from whole-genome sequencing, and their association with human disease. A maximum- likelihood phylogenetic tree was constructed in RaxML (LG+Γ+F model) [111] from the concatenated amino acid sequences derived from 78 near-universal single-copy genes [228]. Support values >50 (from 100 bootstrap replicates) are given at the corresponding nodes. Bar, 0.1 amino acid substitutions per site. Legionella species coloured in red have been associated with human disease, and those in black have not (yet) been linked to disease. Appearing at the top of the list are non-Legionella species (blue) that belong to other genera within the Legionellales. (b) A depiction of the distribution of the 12 core lsp T2SS genes (represented by coloured arrows as in Fig. 1) throughout the order Legionellales. Distinct genetic loci are separated by double slashes. White arrows indicate genes unrelated to the Lsp T2SS. Arrows filled with hatch marks indicate pseudogenes. Gene arrows are drawn to scale.

16 White and Cianciotto, Microbial Genomics 2019;5

Fig. 6. Chromosomal organization of the T2SS genes in different Legionella species. The genetic context of the fivelsp gene clusters within 12 fully sequenced Legionella species was determined using SimpleSynteny [230]. Dark grey arrows depict lsp genes. Other coloured arrows represent genes flanking the lsp gene clusters. Orthologous genes are joined by vertical lines. The genomic coordinates are given beneath each genome segment. that extensive chromosomal rearrangements have occurred (N=255) are found in only nine or fewer of the Legionella throughout the evolution of Legionella species, as has been species (Fig. 7a) [196, 229]. Moreover, seven out of the 72 previously described [217, 224]. T2SS (documented+putative) effectors (9.7 %) are conserved in all 57 species and thereby represent ‘core’ effectors. Once Distribution of T2SS substrates again, this level of conservation is rather different from the across the genus Legionella T4SS where there are only eight core effectors out of 255 T4SS effectors examined (3.1 %) [196, 229]. There appears to be General patterns and distributions for specific only one L. pneumophila-specific T2SS effector, namely the substrates putative effector Lpg0165 (Table S2). This presents yet another The 25 T2SS substrates that have been confirmed for distinction from the Dot/Icm system, where at least 20 T4SS L. pneumophila strain 130b exhibit a range of distributions effectors are L. pneumophila-specific [229]. In summary, a across the genus Legionella (Table 1), reinforcing an earlier majority of the T2SS effectors appear to be shared by a large conclusion that was based on the prevalence of LapA, NttD, subset of Legionella species, hinting at a critical role for them PlaC and ProA within Legionella [75]. Extending this analysis in the ecology of Legionella owing to their long evolutionary to include both the 25 validated and the 47 putative T2SS history across the genus. substrates, it appears that the vast majority of T2SS effectors associated with L. pneumophila are found in 32 or more Turning attention specifically to the distribution of the 25 of the 57 Legionella species analysed (Fig. 7a). This stands known T2SS effectors (Table 1), the metalloprotease ProA and in marked contrast to the situation for the Dot/Icm T4SS, the phospholipase A/lysophospholipase A PlaA are notable for where the majority of a subset of T4SS effectors analysed being found in all 57 species analysed, and thus constitute the

17 White and Cianciotto, Microbial Genomics 2019;5

Fig. 7. The genus-wide prevalence of L. pneumophila effectors. (a) The distribution of documented Lsp T2SS effectors (N=25), putative T2SS effectors (N=47), and a subset of documented Dot/Icm T4SS effectors N( =255) among the 57 analysed Legionella species was determined. The relative frequency of effector groups (y-axis) was determined by the number of species genomes in which individual effectors were present (x-axis), with L. pneumophila-specific effectors at the far left (i.e. x=1) and core effectors at the far right (i.e. x=57). (b) The role of validated T2SS effectors in protozoan infection versus the genus-wide prevalence was determined for N=24 experimentally characterized T2SS effectors based on mutant analysis in a protozoan infection model. Open symbols represent those effectors for which genetic complementation has not yet been achieved. A Student’s t-test was performed between the two sample distributions. The dashed line represents the median genus-wide prevalence for the analysed effectors. first examples of ‘core’ effectors of the Legionella T2SS (Fig. 8). protozoa V. vermiformis and A. castellanii but not intracellular While only ProA and PlaA are found within all genomes, growth per se [116]. Overall, this analysis indicates that the the acylglycerol lipase LipA, aminopeptidase LapA, novel T2SS effectors known to be involved in intracellular infection effector NttF and ribonuclease SrnA were found in 89–95 % of at least one eukaryotic host are significantly more prevalent of the Legionella species (Fig. 8). Interestingly, LipA and ProA throughout Legionella as compared to those effectors that are are found immediately adjacent to one another within the not required for intracellular infection of natural host cells L. pneumophila chromosome (Fig. 3), and the rate of (Fig. 7b). Thus, we hypothesize that T2SS effectors that are co-occurrence and synteny in the genome was 89 % among the more prevalent within the genus Legionella are under stronger 57 Legionella species analysed. Since effector genes encoded selective pressure due to their role in infection of the natural within close proximity in the L. pneumophila genome may host. coordinate their functions or regulate one another [231, 232], ProA and LipA might function in a coordinated fashion. Six Groupings amongst the Legionella species based on other effectors, i.e. phospholipase A PlaC, putative amidase their T2SS substrates AmiA, and the novel effectors NttA, NttB, NttC and NttD, were found in 75–86 % of species (Fig. 8). Of the 12 effectors Although most of the documented T2SS effectors are widely with >75 % conservation, seven (i.e. LapA, NttA, NttC, NttD, distributed across the genus Legionella, no other Legionella PlaC, ProA and SrnA) clearly promote infection of at least one species analysed possessed the same effector repertoire protozoan host [25, 76, 113, 158]. Additionally, preliminary that was found in L. pneumophila (Fig. 8). L. quateirensis, studies suggest that AmiA and novel effector NttF may also L. fallonii and L. waltersii were most similar to L. pneumophila promote infection of protozoa [140, 153], potentially bringing in this regard with each having 21 or 22 of the effectors (Fig. 8). the total to nine out of 12. Although plaA mutants have thus L. longbeachae, the second most common cause of Legion- far not been shown to be impaired in infection, PlaA appears naires’ disease, has 18 of the effectors (Fig. 8). L. geestiana, to influence the integrity of the LCV membrane, with a mutant L. fairfieldensis, L. londiniensis, L. adelaidensis, L. impletisoli lacking the T4SS effector SdhA producing a highly unstable and L. yabuuchiae, the species most distantly related to LCV, and this loss of LCV membrane integrity was reversed L. pneumophila, possessed the lowest number of shared effec- upon subsequent mutation of PlaA [156]. Thus, the vast tors at 7–10 (Fig. 8). Some of the species examined had shared majority of the known T2SS effectors that are highly prevalent subsets of the T2SS substrates, and there appeared to be six within the genus are implicated in L. pneumophila infection distinct groupings of species based upon these shared effec- of amoebae. In contrast, nine T2SS effectors (i.e. CelA, ChiA, tors. The first group included L. wadsworthii, L. steigerwaltii, GamA, LegP, LipB, Map, NttE, NttG, PlcB) have a prevalence L. anisa, L. bozemanii, L. parisiensis and L. steelei, which lack of 33–74 % within the Legionella species (Fig. 8), and none of LapB, Lcl, LegP, LirB, PlcA and PlcB (green cells in Fig. 8). them are yet clearly implicated in infection (Table 1). Finally, While this effector repertoire is found among phylogeneti- four effectors, i.e. LapB, Lcl, LirB and PlcA, were found in cally related Legionella species, other effector repertoires are fewer than 20 % of Legionella species (Fig. 8). LapB and PlcA interspersed within this clade. Thus, perhaps the repertoire are dispensable for infection of protozoa [25, 143], as noted shapes the environmental niche, and is not simply defined above, whereas Lcl promotes attachment and invasion in the by the evolution of the genus. The second group included

18 White and Cianciotto, Microbial Genomics 2019;5

Fig. 8. Distribution of T2SS substrates across the genus Legionella and beyond. The presence/absence of the 25 Lsp T2SS substrates in all sequenced members of the order Legionellales was determined using blastp as previously described [76]. Black cells indicate the presence of all 25 substrates in the L. pneumophila genome. Rows of the same colour (with the exception of dark blue) indicate effector repertoires shared by more than one Legionella species. Selected clades undergoing effector gain/loss are highlighted in grey and numbered. Bar, 0.1 amino acid substitutions per site.

L. cherrii, L. dumoffii and L. gormanii, which lack LapB, Lcl, thus, these species may also occupy another environmental LegP, LirB, Lpw18641, PlcA and PlcB (purple cells in Fig. 8). niche. The third group included L. gratiana, L. cincinnatiensis, Like the first repertoire, L. cherrii and L. dumoffii are paraphy- L. santicrucis and L. longbeachae, which lack ChiA, LapB, Lcl, letic, while L. gormanii belongs to a different phyletic group; LegP, LirB, NttB and PlcA (gold cells in Fig. 8). All four of these

19 White and Cianciotto, Microbial Genomics 2019;5

Fig. 9. Phylogenetic analysis of Lsp T2SS proteins of Legionella. Homologes of Legionella T2SS apparatus proteins were identified by blastp using a minimum query coverage of 60 % for T2S DEFGHIJKLMO or 30 % for T2S C, and amino acid identity and E-value cutoffs of 30 and 1×10−10, respectively, for T2S DEFGIO, and amino acid identity and E-value cutoffs of 20 and 1×10−5, respectively, for T2S CHJKLM. Maximum-likelihood trees were generated from full-length amino acid alignments of all 12 individual T2SS apparatus proteins (i.e. T2S CDEFGHIJKLMO) and the most closely related homologues (as determined by blastp) encompassing 20 unique genera using RaxML (100 bootstrap replicates, GTR + Γ model) [111]. Bootstrap support values >50 are presented at the respective nodes. Bar, 0.1 amino acid substitutions per site. Labels representing Legionella Lsp proteins are in bold green, and labels representing Aquicella proteins are in bold purple. Monophyletic clades containing only Legionella and Aquicella T2SS orthologues are shaded in blue.

20 White and Cianciotto, Microbial Genomics 2019;5

species are paraphyletic, and thus this repertoire probably [76]. The first substrate, PlaC, appears to be an ‘ancestral’ T2SS arose from divergent evolution (i.e. gene gain and loss) within effector that has undergone two loss events over time, once a single Legionella clade. The fourth group was the monophy- in clade ‘I’ and once in clade ‘II’ (Fig. 8, grey shaded regions). letic group of L. rubrilucens, L. erythra and L. tauriniensis, The NttD substrate is another ancestral T2SS effector, but one which lack NttA, LipB, Map, CelA, PlcA, LapB, LirB and Lcl that seems to have undergone three loss events: once in clade I (light blue cells in Fig. 8). The fifth was the monophyletic (Fig. 8, grey shaded region), once in L. fairfieldensis and once group of L. jamestowniensis and L. clemsonensis, which lacked in L. drozanskii. LapA appears to be a second ancestral T2SS Lpw_02811 (Lpg0189), LipB, Map, CelA, PlcA, PlcB, LapB, effector that has undergone three loss events: once in L. naga- LirB and Lcl (beige cells in Fig. 8). The final group was the sakiensis, once in L. londiniensis and once in L. maceachernii. monophyletic group of L. nautarum and L. drozanskii, which On the other hand, LapB probably arose from a recent gene lacked NttC, PlaC, GamA, NttE, NttG, ChiA, Map, LegP, duplication of LapA, having occurred twice: once in clade ‘III’ CelA, PlcA, PlcB, LirB, LapB and Lcl (grey cells in Fig. 8). All (and subsequently lost in L. norrlandica) and once in clade other Legionella species possess ‘unique’ repertoires, which, ‘IV’ (Fig. 8, grey shaded regions). This second gene copy prob- overall, range in size from 25 in L. pneumophila (black cells) to ably underwent positive selection and emerged with a new seven in L. yabuuchiae. Overall, this comparison of the T2SS substrate specificity that is non-redundant with the closely effector repertoires across the 57 examinedLegionella species related LapA [76, 143]. suggests that phylogenetically more closely related species share similar sets of effectors. However, this L. pneumophila- centric view of T2SS effectors may underestimate the true On the origins of the Legionella Lsp number and distributions of Legionella T2SS effectors, as T2SS there are T2SS-compatible extracellular activities that are not dependent on the 25 known L. pneumophila effectors, Legionella species (the sole members of the family Legionel- as noted above. Furthermore, it is entirely possible that there laceae) are most closely related to members of the family are T2SS effectors expressed by non-pneumophila species which contains , , Aquicella, that are absent from L. pneumophila. In this vein, it is worth Berkiella, Occultobacter and Nucleophilum among others, noting that there are >600 orthologous T4SS effectors across which together with Legionellaceae make up the order the genus Legionella, approximately half of which are absent Legionellales [3]. It was previously reported that Coxiella from L. pneumophila [228]. burnettii, although possessing a T4SS, lacks a core set of T2SS genes [50, 237]. Therefore, it had remained unclear At least 15 examined species of Legionella beyond L. pneu- when the Legionella T2SS emerged within the Legionellales, mophila are known to possess secreted activities compatible if at all within the closely related Coxiellaceae. When the with T2S. Given that homologues of ProA are found among additionally available Coxiellaceae genomes encompassing all sequenced Legionella genomes, it is not surprising that three species of Rickettsiella [RefSeq assembly accessions L. anisa, L. cincinnatiensis, L. dumoffii, L. erythra, L. feeleii, GCF_001881485.1 (Rickettsiella grylli), GCF_000168295.1 L. gormanii, L. jordanis, L. longbeachae, L. moravica, L. paris- (R. grylli), GCF_003966755.1 (Rickettsiella viridis) and iensis, L. steigerwaltii and L. wadsworthii all secrete protease GCF_001881495.1 (Rickettsiella isopodorum)], one species of activity [41, 233, 234]. In a similar way, it is logical that a phos- Diplorickettsia (RefSeq assembly accession GCF_000257395.1 pholipase A activity, robably due to PlaA, has been detected (Diplorickettsia massiliensis)], and two of Berkiella [RefSeq in supernatants from L. anisa, L. dumoffii, L. gormanii, assembly accessions GCF_001431295.1 (Berkiella aquae) and L. jordanis, L. longbeachae, L. oakridgensis, L. parisiensis and GCF_001431315.1 (Berkiella cookevillensis)] were examined, L. steigerwaltii [233, 235]. Moreover, L. dumoffii, L. gormanii there was also no evidence of T2SS apparatus genes, other and L. steigerwaltii supernatants are positive for acid phos- than pilD/lspO, which was linked, as is often the case, to other phatase activity, compatible with their genomes’ encoding T4P genes. However, further analysis revealed a complete homologues of Map (Fig. 8) [233]. L. longbeachae, L. boze- set of T2SS components within the genome of the very manii, L. dumoffii, L. gormanii, L. jordanis and L. micdadei recently sequenced A. lusitana (RefSeq assembly accession all possess secreted phospholipase C activity [236]. However, GCF_003350455.1), an aquatic bacterium within the Coxiel- only L. longbeachae encodes a homologue of either PlcA or laceae [196]. Moreover, five of 12 core Lsp proteins from PlcB, and although L. erythra secretes an endoglucanase L. pneumophila (i.e. LspD, LspF, LspI, LspK and LspL) shared activity [141], it lacks a homologue of CelA (Fig. 8). These their highest amino-acid identity with the orthologous T2SS latter observations lend support to the view that there are components of A. lusitana, with identity ranging from 24 % additional T2SS effectors within the genus Legionella that are for LspL to 49 % for LspF (Fig. 9). All of the other seven Lsp not present within L. pneumophila. proteins also showed strong sequence relatedness to their Aquicella counterparts, although their closest homologues Examples of how T2SS substrates are gained and existed in various other types of Gammaproteobacteria. lost within the genus Phylogenetic analysis provided further evidence that the Based on the distribution patterns summarized above, a majority of Lsp proteins (10 of 12) are most closely related recent study proposed possible scenarios by which select T2SS to proteins in A. lusitana. (Fig. 9). In the case of LspC, due to substrates were gained and lost within the genus Legionella the high sequence divergence among related T2S C proteins,

21 White and Cianciotto, Microbial Genomics 2019;5 the evolutionary history could not be reliably inferred, with Concluding thoughts very few branches containing bootstrap values >50. In the That the L. pneumophila T2SS, with its 25 validated substrates case of PilD, the branch was adjacent to that of A. lusitana (Table 1), has a major role in the ecology and pathogenesis T2S O; however, it was not monophyletic but intermediate of the Legionnaires’ disease agent is now well known, as between a clade containing Aquicella and a clade containing summarized in the initial sections of this review. Looking Spongiibacter, which has also been isolated from protozoa to the future, it will be instructive to confirm whether [238]. Given the dual role of T2S O in both protein secretion the many putative substrates of the L. pneumophila T2SS and T4P biogenesis, and that T4P are present in all members (Table S2) are bona fide substrates. Such a finding would of the order Legionellales, the evolutionary trajectory of T2S clearly document that the output of a T2SS can be quite large O is not as clear. Therefore, it appears that the T2SS appa- and varied, perhaps encompassing a wealth of novel enzymes. ratus of Legionella is closer to that of Aquicella than to any In the meantime, it will be important to more precisely define other bacterial genus. Whereas the environmental reservoirs the enzymatic activities and molecular modes of action of the of the obligate intracellular bacteria belonging to Coxiella known T2SS substrates, particularly those that are required and Rickettsiella/Diplorickettsia are thought to be primarily for the ability of L. pneumophila to infect host cells, evade mammals and arthropods, respectively [239–241], Aquicella immune defences or mediate damage to tissue. Happily, like Legionella can be routinely cultured in the laboratory and substrates of the Legionella T2SS have recently garnered the replicates intracellularly within aquatic protozoa [196, 242]. attention of structural biologists leading to the reporting of Interestingly, Berkiella species are the closest relative to nine crystal structures (Tables 1 and S2). Further expansion Legionella yet are obligate intracellular parasites of amoebae in this dataset will probably enhance both our understanding and replicate inside the host cell nucleus [243]. Thus, we of substrate activities and the mechanism of the secretion posit two scenarios for the emergence of the Legionella process itself, including elucidating how the 3-D structures Lsp T2SS: the Lsp-like T2SS apparatus emerged within the of the substrates are recognized by the secretion apparatus. Legionellales in a common ancestor shared between Rickett- Another fruitful area for future investigation is delving siella–Diplorickettsia–Aquicella–Berkiella–Legionella, and was more deeply into the regulatory networks that control the lost twice, once in the Rickettsiella–Diplorickettsia clade and expression of the T2SS apparatus and/or its many different once in the Berkiella clade; alternatively, the Lsp-like T2SS effectors. The available data (Table S3) already indicate that apparatus emerged within A. lusitana, and was subsequently the regulation of T2SS function is complex and multifacto- acquired in a Legionella progenitor via HGT within protozoa. rial; nonetheless, deciphering how the activity of the T2SS is While it is unknown whether the T2SS contributes to the coordinated with that of the Dot/Icm T4SS and other systems ability of Aquicella to replicate within the protozoa (in the is critical for understanding the overall virulence strategy of cytosol, or at minimum not intranuclear) and outside of the the Legionella . host, we posit that this is the case based upon the compel- ling importance that T2SS has in the intracellular parasitism The rapidly expanding number of genome sequences available and extracellular survival and persistence of L. pneumophila. for the genus Legionella and beyond has greatly facilitated our Intriguingly, the A. lusitana genome possesses homologues understanding of the distribution and diverse origins of the of six of the 25 known L. pneumophila T2SS effectors, T2SS and its arsenal of effectors, as presented in the latter part including ProA (E=1×10−143), PlaA (E=2.16×10−45), GamA of this review. It is now clear that the genes encoding the T2SS (E=5.3×10−65), LapA (E=1.0×10−113), PlcA (E=3.20×10−136) apparatus are absolutely conserved across the genus Legionella, and ChiA (E=1.12x10−150), a pattern that is not too dissimilar which includes 62 species and more than 30 pathogens in from that of L. londiniensis and L. adelaidensis, both of which addition to L. pneumophila. Moreover, the vast majority of have nine out of the 25 (Fig. 8). The fact that at least ProA, the T2SS effectors associated with L. pneumophila are shared PlaA and LapA have a role in intracellular infection by by a large number of other Legionella species, signalling at L. pneumophila also further implies an importance for T2SS a key role for them in the ecology of Legionella as a whole. in the intracellular parasitism and ecology of Aquicella. Inter- However, no other species has the same effector repertoire as estingly, five of the L. pneumophila effectors had their closest does L. pneumophila, with, as a general rule, phylogenetically homologue occurring in A. lusitana, more than for any other more closely related Legionella species sharing similar sets of effectors. Interestingly, the T2SS effectors that are involved in genus (Table 1). In contrast, current blastp analysis found no intracellular infection of a eukaryotic host(s) are significantly orthologues to any of the known T2SS effectors ofLegionella more prevalent throughout Legionella, indicating that they are in C. burnetii, D. massiliensis and R. viridis, and only a couple under stronger selective pressure. in species of other Rickettsiella or Berkiella genomes (Fig. 8). Given the complete lack of T2SS genes in C. burnetii, we Based on these genomic data, we can also posit a scenario by hypothesize that the T2SS emerged after the divergence of which the L. pneumophila T2SS evolved (Fig. 10). To begin, Coxiella species from other Coxiellaceae (and Legionellaceae) it is hypothesized that the T2SS emerged within a common members. In summary, based upon the latest updates in the ancestor of Aquicella, Berkiella and Legionella, helping to genome database, we suggest that the T2SS of L. pneumophila promote an intra-amoebal lifestyle. The T2SS was lost within originated from within the order Legionellales and that many Berkiella species, and perhaps this event had some connec- of the effectors may have also arisen within that progenitor. tion to the Berkiellae becoming obligate intracellular parasites,

22 White and Cianciotto, Microbial Genomics 2019;5

Fig. 10. Model for evolution of T2SS and its effectors within the genus Legionella. The acquisition of T2SS effectors over (evolutionary) time, with the last common ancestor among Legionella and Aquicella at the left, and L. pneumophila at the lower right. Genes are indicated by coloured rectangles, and HGT is indicated by the curved arrows. The divergence of Aquicella and the non-pneumophila Legionella species is indicated with vertical, dashed arrows. L. yabuuchiae is an example of a species that shares 7–10 of 25 effectors with L. pneumophila; L. longbeachae shares 18/25 effectors; and L. fallonii is representative of species sharing 21 or 22 of 25 effectors. Within the infected amoebae hosts in the centre of the figure, the black circles represent the nucleus of the protozoan hosts, whereas the white circles represent contractile vacuoles. targeting the host cell nucleus for survival. The acquisition of the acquisition of even more T2SS effectors, along with other core effector genes (e.g. proA) probably helped to shape the events, such as the evolution of the Dot/Icm T4SS, led to the early evolution of the Aquicella–Legionella ancestor (Fig. 10, emergence of the L. pneumophila species (Fig. 10, step 6). step 1). With time, Aquicella and Legionella diverged from Based upon the differences in the known-effector repertoire each other (Fig. 10, step a), although both retained their T2SS amongst the Legionella species (Fig. 8), we posit that each of and remained as facultative intracellular parasites of amoebae. the different Legionella species/clades travelled along their The ancestral Legionella appears to have acquired additional own path of T2SS evolution, which probably includes the effectors via inter-kingdom HGT, owing to its natural compe- acquisition of T2SS substrates that do not have homologues tence and ability to incorporate environmental DNA. As one in L. pneumophila (Fig. 10, steps b, c and d). example of HGT, Legionella probably acquired genetic mate- rial from its protozoan host, giving rise to eukaryotic-like In the coming years, we anticipate the discovery of additional T2SS substrates such as LapA (Fig. 10, step 2). We predict L. pneumophila-like T2SSs and new genome sequences that that a number of the T2SS substrates that are currently will provide further insight into the diverse origins of the described as ‘novel’ will be re-classified as eukaryotic-like many effectors in the expansive genusLegionella . Finally, the as more protozoan genomes are sequenced. While growing genomic analysis of the L. pneumophila T2SS that has been within its amoebal hosts, Legionella probably encountered reviewed here can serve as a model for the investigation of giant viruses (mimiviruses) that also parasitize protozoa. other bacterial T2SSs, especially those that are present in This co-habitation may have provided another conduit for aquatic and/or intracellular parasites of protozoa. the HGT of effectors, including the T2SS substrate ChiA (Fig. 10, step 3). When Legionella emerges from its spent protozoan hosts, it encounters a wide variety of other organ- Funding information This work was funded by NIH grant AI043987 awarded to N. P. C. isms in its aquatic environment, such as cyanobacteria, water moulds and red algae. This undoubtedly provided yet addi- Acknowledgements tional opportunities for gene acquisition, accounting for the We thank members of the Cianciotto lab past and present for helpful T2SS substrates GamA, Lcl and LirB, among others (Fig. 10, advice and assistance with some preliminary data. step 4). It is reasonable to think that Legionella’s host range Conflicts of interest grew as its T2SS effector repertoire expanded. Consequently, The authors declare that there are no conflicts of interest. Legionella may have shared ecological niches with other References intracellular bacterial pathogens, such as Francisella species, 1. Fraser DW, Tsai TR, Orenstein W, Parkin WE, Beecham HJ et al. and thereby acquired further effectors, such as Map, NttG and Legionnaires' disease: description of an epidemic of pneumonia. SrnA, via inter-bacterial HGT (Fig. 10, step 5). Ultimately, N Engl J Med 1977;297:1189–1197.

23 White and Cianciotto, Microbial Genomics 2019;5

2. McDade JE, Shepard CC, Fraser DW, Tsai TR, Redus MA et al. 22. Rowbotham TJ. Current views on the relationships between Legionnaires' disease: isolation of a bacterium and demon- amoebae, legionellae and man. Isr J Med Sci 1986;22:678–689. stration of its role in other respiratory disease. N Engl J Med 23. Declerck P, Behets J, Delaedt Y, Margineanu A, Lammertyn E et al. 1977;297:1197–1203. Impact of non-Legionella bacteria on the uptake and intracellular 3. Duron O, Doublet P, Vavre F, Bouchon D. The importance of revis- replication of Legionella pneumophila in Acanthamoeba castellanii iting Legionellales diversity. Trends Parasitol 2018;34:1027–1037. and Naegleria lovaniensis. Microb Ecol 2005;50:536–549. 4. Alary M, Joly JR. Risk factors for contamination of domestic 24. Thomas V, Herrera-Rimann K, Blanc DS, Greub G. Biodiversity hot water systems by legionellae. Appl Environ Microbiol of amoebae and amoeba-resisting bacteria in a hospital water 1991;57:2360–2367. network. Appl Environ Microbiol 2006;72:2428–2438. 5. Fliermans CB, Cherry WB, Orrison LH, Smith SJ, Tison DL et al. 25. Tyson JY, Pearce MM, Vargas P, Bagchi S, Mulhern BJ et al. Ecological distribution of Legionella pneumophila. Appl Environ Multiple Legionella pneumophila type II secretion substrates, Microbiol 1981;41:9–16. including a novel protein, contribute to differential infec- 6. Mouchtouri V, Velonakis E, Tsakalof A, Kapoula C, Goutz- tion of the amoebae Acanthamoeba castellanii, Hartman- iana G et al. Risk factors for contamination of hotel water distri- nella vermiformis, and Naegleria lovaniensis. Infect Immun bution systems by Legionella species. Appl Environ Microbiol 2013;81:1399–1410. 2007;73:1489–1492. 26. Winn WC, Jr. Legionnaires disease: historical perspective. Clin 7. Ashbolt NJ. Environmental (Saprozoic) pathogens of engineered Microbiol Rev 1988;1:60–81. water systems: understanding their ecology for risk assessment 27. Gao LY, Stone BJ, Brieland JK, Abu Kwaik Y. Different fates of and management. Pathogens 2015;4:390–405. Legionella pneumophila PMI and mil mutants within macrophages 8. Palmer A, Painter J, Hassler H, Richards VP, Bruce T et al. and alveolar epithelial cells. Microb Pathog 1998;25:291–306. Legionella clemsonensis sp. nov.: a green fluorescing Legionella 28. Maruta K, Miyamoto H, Hamada T, Ogawa M, Taniguchi H et al. strain from a patient with pneumonia. Microbiol Immunol Entry and intracellular growth of in alveolar 2016;60:694–701. epithelial cells. Am J Respir Crit Care Med 1998;157:1967–1974. 9. Relich RF, Schmitt BH, Raposo H, Barker L, Blosser SJ et al. 29. Isberg RR, O'Connor TJ, Heidtman M. The Legionella pneumophila Legionella indianapolisensis sp. nov., isolated from a patient with replication vacuole: making a cosy niche inside host cells. Nat pulmonary abscess. Int J Infect Dis 2018;69:26–28. Rev Microbiol 2009;7:13–24. 10. Campocasso A, Boughalmi M, Fournous G, Raoult D, La Scola B. 30. Finsel I, Hilbi H. Formation of a pathogen vacuole according to Legionella tunisiensis sp. nov. and sp. nov., Legionella pneumophila: how to kill one bird with many stones. isolated from environmental water samples. Int J Syst Evol Micro- Cell Microbiol 2015;17:935–950. biol 2012;62:3003–3006. 31. Bandyopadhyay P, Liu S, Gabbai CB, Venitelli Z, Steinman HM. 11. Rizzardi K, Winiecka-Krusnell J, Ramliden M, Alm E, Andersson S Environmental mimics and the Lvh type IVA secretion system et al. Legionella norrlandica sp. nov., isolated from the biopurifica- contribute to virulence-related phenotypes of Legionella pneu- tion systems of wood processing plants. Int J Syst Evol Microbiol mophila. Infect Immun 2007;75:723–735. 2015;65:598–603. 32. Bandyopadhyay P, Lang EAS, Rasaputra KS, Steinman HM. 12. Bajrai LH, Azhar EI, Yasir M, Jardot P, Barrassi L et al. Legionella Implication of the VirD4 coupling protein of the Lvh type 4 secre- saoudiensis sp. nov., isolated from a sewage water sample. Int J tion system in virulence phenotypes of Legionella pneumophila. J Syst Evol Microbiol 2016;66:4367–4371. Bacteriol 2013;195:3468–3475. 13. Ishizaki N, Sogawa K, Inoue H, Agata K, Edagawa A et al. 33. Fuche F, Vianney A, Andrea C, Doublet P, Gilbert C. Functional Legionella thermalis sp. nov., isolated from hot spring water in type 1 secretion system involved in Legionella pneumophila viru- Tokyo, Japan. Microbiol Immunol 2016;60:203–208. lence. J. Bacteriol. 2015;197:563–571. 14. Cianciotto NP, Hilbi H, Buchrieser C. Legionnaires' Disease. The 34. Allard KA, Dao J, Sanjeevaiah P, McCoy-Simandle K, Chatfield CH Prokaryotes - Human Microbiology, 4th edition. New York, NY: et al. Purification of Legiobactin and Importance of This Sidero- Springer; 2013. pp. 147–217. phore in Lung Infection by Legionella pneumophila. Infection and 15. Abu Kwaik Y, Gao LY, Stone BJ, Venkataraman C, Immunity 2009;77:2887–2895. Harb OS. Invasion of protozoa by Legionella pneumophila and its 35. Burnside DM, Wu Y, Shafaie S, Cianciotto NP. The Legionella pneu- role in bacterial ecology and pathogenesis. Appl Environ Microbiol mophila siderophore legiobactin is a polycarboxylate that is iden- 1998;64:3127–3133. tical in structure to rhizoferrin. Infect Immun 2015;83:3937–3945. 16. Adeleke A, Pruckler J, Benson R, Rowbotham T, 36. Chatfield CH, Cianciotto NP. The secreted pyomelanin pigment Halablab M et al. Legionella-like amebal pathogens-phylogenetic of Legionella pneumophila confers ferric reductase activity. Infect status and possible role in respiratory disease. Emerg Infect Dis Immun 2007;75:4062–4070. 1996;2:225–230. 37. Zheng H, Chatfield CH, Liles MR, Cianciotto NP. Secreted 17. Wullings BA, van der Kooij D. Occurrence and genetic diver- pyomelanin of Legionella pneumophila promotes bacterial iron sity of uncultured Legionella spp. in drinking water treated uptake and growth under iron-limiting conditions. Infect Immun at temperatures below 15 degrees C. Appl Environ Microbiol 2013;81:4182–4191. 2006;72:157–166. 38. Hales LM, Shuman HA. Legionella pneumophila contains a type 18. Gomaa F, Gersh M, Cavanaugh CM. Diverse Legionella- II general secretion pathway required for growth in amoebae Like bacteria associated with testate amoebae of the genus as well as for secretion of the Msp protease. Infect Immun Arcella (Arcellinida: Amoebozoa). J Eukaryot Microbiol 1999;67:3662–3666. 2018;65:661–668. 39. Rossier O, Cianciotto NP. Type II protein secretion is a subset of 19. Diederen BMW. Legionella spp. and Legionnaires' disease. J Infect the PilD-dependent processes that facilitate intracellular infec- 2008;56:1–12. tion by Legionella pneumophila. Infect Immun 2001;69:2092–2098. 20. Boamah DK, Zhou G, Ensminger AW, O'Connor TJ, Hosts FM. 40. Liles MR, Edelstein PH, Cianciotto NP. The prepilin peptidase From many hosts, one accidental pathogen: the diverse proto- is required for protein secretion by and the virulence of the zoan hosts of Legionella. Front Cell Infect Microbiol 2017;7:477. intracellular pathogen Legionella pneumophila. Mol Microbiol 21. Anand CM, Skinner AR, Malic A, Kurtz JB. Interaction of L. pneu- 1999;31:959–970. mophilia and a free living amoeba (Acanthamoeba palestinensis). 41. Söderberg MA, Dao J, Starkenburg SR, Cianciotto NP. Impor- J Hyg 1983;91:167–178. tance of type II secretion for survival of Legionella pneumophila

24 White and Cianciotto, Microbial Genomics 2019;5

in tap water and in amoebae at low temperatures. Appl Environ ATPase-pseudopilin coupling protein in the Type II secretion Microbiol 2008;74:5583–5588. system of . Mol Microbiol 2011;79:786–798. 42. Rossier O, Starkenburg SR, Cianciotto NP. Legionella pneu- 63. Peabody CR, Chung YJ, Yen MR, Vidal-Ingigliardi D, mophila type II protein secretion promotes virulence in the A/J Pugsley AP et al. Type II protein secretion and its relation- mouse model of Legionnaires' disease pneumonia. Infect Immun ship to bacterial type IV pili and archaeal flagella. Microbiology 2004;72:310–321. 2003;149:3051–3072. 43. Cianciotto NP, White RC. Expanding role of type II secretion in 64. Korotkov KV, Sandkvist M, Hol WGJ. The type II secretion system: bacterial pathogenesis and beyond. Infect Immun 2017;85. biogenesis, molecular architecture and mechanism. Nat Rev Microbiol 2012;10:336–351. 44. Gu S, Shevchik VE, Shaw R, Pickersgill RW, Garnett JA. The role of intrinsic disorder and dynamics in the assembly and func- 65. Wang X, Pineau C, Gu S, Guschinskaya N, Pickersgill RW et al. tion of the type II secretion system. Biochimica et biophysica acta Cysteine scanning mutagenesis and disulfide mapping analysis 1865;2017:1255–1266. of arrangement of GspC and GspD protomers within the type 2 secretion system. J Biol Chem 2012;287:19082–19093. 45. Rondelet A, Condemine G. Type II secretion: the substrates that won't go away. Res Microbiol 2013;164:556–561. 66. Gérard-Vincent M, Robert V, Ball G, Bleves S, Michel GPF et al. Identification of XcpP domains that confer functionality and 46. d'Enfert C, Reyss I, Wandersman C, Pugsley AP. Protein secretion specificity to the Pseudomonas aeruginosa type II secretion appa- by gram-negative bacteria. Characterization of two membrane ratus. Mol Microbiol 2002;44:1651–1665. proteins required for pullulanase secretion by K-12. J Biol Chem 1989;264:17462–17468. 67. Pineau C, Guschinskaya N, Robert X, Gouet P, Ballut L et al. Substrate recognition by the bacterial type II secretion system: 47. d'Enfert C, Ryter A, Pugsley AP. Cloning and expression in more than a simple interaction. Mol Microbiol 2014;94:126–140. Escherichia coli of the Klebsiella pneumoniae genes for produc- tion, surface localization and secretion of the lipoprotein pullula- 68. Douzi B, Ball G, Cambillau C, Tegoni M, Voulhoux R. Deciphering nase. Embo J 1987;6:3531–3538. the Xcp Pseudomonas aeruginosa type II secretion machinery through multiple interactions with substrates. J Biol Chem 48. Filloux A. The underlying mechanisms of type II protein secre- 2011;286:40792–40801. tion. Biochim Biophys Acta 2004;1694:163–179. 69. Bouley J, Condemine G, Shevchik VE. The PDZ domain of OutC 49. Voulhoux R, Ball G, Ize B, Vasil ML, Lazdunski A et al. Involvement and the N-terminal region of OutD determine the secretion speci- of the twin-arginine translocation system in protein secretion via ficity of the type II out pathway of Erwinia chrysanthemi. J Mol Biol the type II pathway. EMBO J 2001;20:6735–6741. 2001;308:205–219. 50. Cianciotto NP. Type II secretion: a protein secretion system for all 70. Michel-Souzy S, Douzi B, Cadoret F, Raynaud C, Quinton L et al. seasons. Trends Microbiol 2005;13:581–588. Direct interactions between the secreted effector and the T2SS 51. Cianciotto NP. Many substrates and functions of type II secretion: components GspL and GspM reveal a new effector-sensing step lessons learned from Legionella pneumophila. Future Microbiol during type 2 secretion. J Biol Chem 2018;293:19441–19450. 2009;4:797–805. 71. Korotkov KV, Krumm B, Bagdasarian M, Hol WGJ. Struc- 52. Korotkov KV, Sandkvist M. Architecture, function, and substrates tural and functional studies of EpsC, a crucial component of of the type II secretion system. EcoSal Plus 2019;8. the type 2 secretion system from Vibrio cholerae. J Mol Biol 2006;363:311–321. 53. Abendroth J, Rice AE, McLuskey K, Bagdasarian M, Hol WGJ. The crystal structure of the periplasmic domain of the type II 72. Bleves S, Gérard-Vincent M, Lazdunski A, Filloux A. Structure- secretion system protein EpsM from Vibrio cholerae: the simplest function analysis of XcpP, a component involved in general version of the ferredoxin fold. J Mol Biol 2004;338:585–596. secretory pathway-dependent protein secretion in Pseudomonas aeruginosa. J Bacteriol 1999;181:4012–4019. 54. Abendroth J, Kreger AC, Hol WGJ. The dimer formed by the peri- plasmic domain of EpsL from the Type 2 secretion system of 73. Sahr T, Rusniok C, Dervins-Ravault D, Sismeiro O, Coppee J-Y Vibrio parahaemolyticus. J Struct Biol 2009;168:313–322. et al. Deep sequencing defines the transcriptional map of L. pneumophila and identifies growth phase-dependent regulated 55. Chen YL, Hu NT. Function-related positioning of the type II secre- ncRNAs implicated in virulence. RNA Biol 2012;9:503–519. tion ATPase of Xanthomonas campestris pv. campestris. PLoS One 2013;8:e59123. 74. Guilvout I, Chami M, Engel A, Pugsley AP, Bayan N. Bacte- rial outer membrane secretin PulD assembles and inserts 56. Lu C, Korotkov KV, Hol WGJ. Crystal structure of the full-length into the inner membrane in the absence of its pilotin. EMBO J ATPase GspE from the type II secretion system 2006;25:5241–5249. in complex with the cytoplasmic domain of GspL. J Struct Biol 2014;187:223–235. 75. Finn RD, Clements J, Eddy SR. HMMER web server: inter- active sequence similarity searching. Nucleic Acids Res 57. Camberg JL, Johnson TL, Patrick M, Abendroth J, Hol WGJ et al. 2011;39:W29–W37. Synergistic stimulation of EpsE ATP hydrolysis by EpsL and 76. White RC, Gunderson FF, Tyson JY, Richardson KH, Portlock TJ acidic phospholipids. EMBO J 2007;26:19–27. et al. Type II secretion-dependent Aminopeptidase LapA and acyl- 58. Hay ID, Belousoff MJ, Lithgow T. Structural basis of type 2 secre- transferase PlaC Are redundant for nutrient acquisition during tion system engagement between the inner and outer bacterial Legionella pneumophila intracellular infection of . MBio membranes. MBio 2017;8:e01344–17. 2018;9:e00528-18. 59. Yan Z, Yin M, Xu D, Zhu Y, Li X. Structural insights into the secretin 77. Dunstan RA, Heinz E, Wijeyewickrema LC, Pike RN, Purcell AW translocation channel in the type II secretion system. Nat Struct et al. Assembly of the type II secretion system such as found in Mol Biol 2017;24:177–183. Vibrio cholerae depends on the novel Pilotin AspS. PLoS Pathog 60. Cisneros DA, Bond PJ, Pugsley AP, Campos M, Francetic O. Minor 2013;9:e1003117. pseudopilin self-assembly primes type II secretion pseudopilus 78. Viarre V, Cascales E, Ball G, Michel GPF, Filloux A et al. HxcQ elongation. EMBO J 2012;31:1041–1053. liposecretin is self-piloted to the outer membrane by its 61. Douzi B, Durand E, Bernard C, Alphonse S, Cambillau C et al. The N-terminal lipid anchor. J Biol Chem 2009;284:33815–33823. XcpV/GspI pseudopilin has a central role in the assembly of a 79. Madan Babu M, Sankaran K. DOLOP-database of bacterial lipo- quaternary complex within the T2SS pseudopilus. J Biol Chem proteins. Bioinformatics 2002;18:641–643. 2009;284:34580–34589. 80. Carter T, Buensuceso RNC, Tammam S, Lamers RP, Harvey H 62. Gray MD, Bagdasarian M, Hol WGJ, Sandkvist M. In vivo et al. The type IVA pilus machinery is recruited to sites of future cross-linking of EpsG to EpsL suggests a role for EpsL as an cell division. MBio 2017;8.

25 White and Cianciotto, Microbial Genomics 2019;5

81. Yahashiri A, Jorgenson MA, Weiss DS. The SPOR domain, a widely 101. DeShazer D, Brett PJ, Burtnick MN, Woods DE. Molecular charac- conserved peptidoglycan binding domain that targets proteins to terization of genetic loci required for secretion of exoproducts in the site of cell division. J Bacteriol 2017;199. Burkholderia pseudomallei. J Bacteriol 1999;181:4661–4664. 82. Douzi B, Filloux A, Voulhoux R. On the path to uncover the bacte- 102. Iwobi A, Heesemann J, Garcia E, Igwe E, Noelting C et al. rial type II secretion system. Philos Trans R Soc Lond B Biol Sci Novel virulence-associated type II secretion system unique 2012;367:1059–1072. to high-pathogenicity Yersinia enterocolitica. Infect Immun 83. Possot OM, Vignon G, Bomchil N, Ebel F, Pugsley AP. Multiple 2003;71:1872–1879. interactions between pullulanase secreton components involved 103. Lee HM, Chen JR, Lee HL, Leu WM, Chen LY et al. Functional in stabilization and cytoplasmic membrane association of PulE. J dissection of the XpsN (GspC) protein of the Xanthomonas camp- Bacteriol 2000;182:2142–2152. estris pv. campestris type II secretion machinery. J Bacteriol 84. Schoenhofen IC, Stratilo C, Howard SP. An ExeAB complex in the 2004;186:2946–2955. type II secretion pathway of Aeromonas hydrophila: effect of ATP- 104. do Vale A, Pereira C, Osorio RC, dos Santos NMS. The Apopto- binding cassette mutations on complex formation and function. genic Toxin AIP56 Is Secreted by the Type II Secretion System of Mol Microbiol 1998;29:1237–1247. Photobacterium damselae subsp. piscicida. Toxins 2017;9:368. 85. Ayers M, Howell PL, Burrows LL. Architecture of the type 105. Waack U, Warnock M, Yee A, Huttinger Z, Smith S et al. CpaA Is II secretion and type IV pilus machineries. Future Microbiol a Glycan-Specific Adamalysin-like Protease Secreted by Acine- 2010;5:1203–1218. tobacter baumannii That Inactivates Coagulation Factor XII. MBio 86. Cazalet C, Rusniok C, Brüggemann H, Zidane N, Magnier A et al. 2018;9:e01606-18. Evidence in the Legionella pneumophila genome for exploita- 106. Carda-Diéguez M, Silva-Hernández FX, Hubbard TP, Chao MC, tion of host cell functions and high genome plasticity. Nat Genet Waldor MK et al. Comprehensive identification of Vibrio vulnificus 2004;36:1165–1173. genes required for growth in human serum. Virulence 2018;9:981–993. 87. Chien M, Morozova I, Shi S, Sheng H, Chen J et al. The genomic sequence of the accidental pathogen Legionella pneumophila. 107. Elhosseiny NM, Elhezawy NB, Attia AS. Comparative proteomics Science 2004;305:1966–1968. analyses of Acinetobacter baumannii strains ATCC 17978 and AB5075 reveal the differential role of type II secretion system 88. Glöckner G, Albert-Weissenberger C, Weinmann E, Jacobi S, secretomes in lung colonization and resistance. Schunder E et al. Identification and characterization of a new Microb Pathog 2019;128:20–27. conjugation/type IVA secretion system (trb/tra) of Legionella pneumophila Corby localized on two mobile genomic islands. Int 108. Saint-Criq V, Villeret B, Bastaert F, Kheir S, Hatton A et al. Pseu- J Med Microbiol 2008;298:411–428. domonas aeruginosa LasB protease impairs innate immunity in mice and humans by targeting a lung epithelial cystic fibrosis 89. Schroeder GN, Petty NK, Mousnier A, Harding CR, Vogrin AJ et al. transmembrane regulator-IL-6-antimicrobial-repair pathway. Legionella pneumophila strain 130b possesses a unique combi- Thorax 2018;73:49–61. nation of type IV secretion systems and novel Dot/Icm secretion system effector proteins. J Bacteriol 2010;192:6001–6016. 109. Jang KK, Lee ZW, Kim B, Jung YH, Han HJ et al. Identifica- tion and characterization of Vibrio vulnificus plpA encoding 90. D'Auria G, Jiménez-Hernández N, Peris-Bondia F, Moya A, a phospholipase A2 essential for pathogenesis. J Biol Chem Latorre A. Legionella pneumophila pangenome reveals strain- 2017;292:17129–17143. specific virulence factors. BMC Genomics 2010;11:181. 110. Cole JR, Wang Q, Fish JA, Chai B, McGarrell DM et al. Ribosomal 91. Sandkvist M. Type II secretion and pathogenesis. Infect Immun database project: data and tools for high throughput rRNA anal- 2001;69:3523–3535. ysis. Nucleic Acids Res 2014;42:D633–D642. 92. Karlyshev AV, MacIntyre S. Cloning and study of the genetic 111. Stamatakis A. RAxML version 8: a tool for phylogenetic anal- organization of the exe gene cluster of Aeromonas salmonicida. ysis and post-analysis of large phylogenies. Bioinformatics Gene 1995;158:77–82. 2014;30:1312–1313. 93. Pugsley AP. The complete general secretory pathway in gram- 112. Polesky AH, Ross JT, Falkow S, Tompkins LS. Identifica- negative bacteria. Microbiol Rev 1993;57:50–108. tion of Legionella pneumophila genes important for infection 94. Francetic O, Pugsley AP. The cryptic general secretory pathway of amoebas by signature-tagged mutagenesis. Infect Immun (gsp) operon of Escherichia coli K-12 encodes functional proteins. 2001;69:977–987. J Bacteriol 1996;178:3544–3549. 113. Tyson JY, Vargas P, Cianciotto NP. The novel Legionella pneu- 95. Abby SS, Cury J, Guglielmini J, Néron B, Touchon M et al. Identifi- mophila type II secretion substrate NttC contributes to infection cation of protein secretion systems in bacterial genomes. Sci Rep of amoebae Hartmannella vermiformis and Willaertia magna. 2016;6:23080. Microbiology 2014;160:2732–2744. 96. Nivaskumar M, Francetic O. Type II secretion system: a magic 114. Söderberg MA, Dao J, Starkenburg SR, Cianciotto NP. Impor- beanstalk or a protein escalator. Biochimica et biophysica acta tance of type II secretion for survival of Legionella pneumophila 1843;2014:1568–1577. in tap water and in amoebae at low temperatures. Appl Environ 97. Snavely EA, Kokes M, Dunn JD, Saka HA, Nguyen BD et al. Reas- Microbiol 2008;74:5583–5588. sessing the role of the secreted protease CPAF in Chlamydia 115. Söderberg MA, Rossier O, Cianciotto NP. The type II protein trachomatis infection through genetic approaches. Pathog Dis secretion system of Legionella pneumophila promotes growth at 2014;71:336–351. low temperatures. J Bacteriol 2004;186:3712–3720. 98. Wang X, Han Q, Chen G, Zhang W, Liu W. A putative Type II secre- 116. Duncan C, Prashar A, So J, Tang P, Low DE et al. Lcl of tion system is involved in cellulose utilization in Cytophaga hutch- Legionella pneumophila is an immunogenic GAG binding isonii. Front Microbiol 2017;8:1482. adhesin that promotes interactions with lung epithelial cells 99. Karaba SM, White RC, Cianciotto NP. Stenotrophomonas malt- and plays a crucial role in biofilm formation. Infect Immun ophilia encodes a type II protein secretion system that promotes 2011;79:2168–2181. detrimental effects on lung epithelial cells. Infect Immun 117. Mallegol J, Duncan C, Prashar A, So J, Low DE et al. Essential 2013;81:3210–3219. roles and regulation of the Legionella pneumophila collagen-like 100. Corbett M, Virtue S, Bell K, Birch P, Burr T et al. Identification adhesin during biofilm formation. PLoS One 2012;7:e46462. of a new quorum-sensing-controlled virulence factor in Erwinia 118. Lucas CE, Brown E, Fields BS. Type IV pili and type II secretion carotovora subsp. atroseptica secreted via the type II targeting play a limited role in Legionella pneumophila biofilm colonization pathway. Mol Plant Microbe Interact 2005;18:334–342. and retention. Microbiology 2006;152:3569–3573.

26 White and Cianciotto, Microbial Genomics 2019;5

119. Stewart CR, Rossier O, Cianciotto NP. Surface translocation by 137. Kulp A, Kuehn MJ. Biological functions and biogenesis of Legionella pneumophila: a form of sliding motility that is dependent secreted bacterial outer membrane vesicles. Annu Rev Microbiol upon type II protein secretion. J Bacteriol 2009;191:1537–1546. 2010;64:163–184. 120. Stewart CR, Burnside DM, Cianciotto NP. The surfactant of 138. Jan AT. Outer membrane vesicles (OMVs) of gram-negative Legionella pneumophila is secreted in a TolC-dependent manner bacteria: a perspective update. Front Microbiol 2017;8:1053. and is antagonistic toward other Legionella species. J Bacteriol 139. Schwechheimer C, Kuehn MJ. Outer-membrane vesicles from 2011;193:5971–5984. gram-negative bacteria: biogenesis and functions. Nat Rev Micro- 121. Johnston CW, Plumb J, Li X, Grinstein S, Magarvey NA. Informatic biol 2015;13:605–619. analysis reveals Legionella as a source of novel natural products. 140. Pearce MM. The Identification and Characterization of a Novel Synth Syst Biotechnol 2016;1:130–136. Clinically Relevent Legionella Species, , and the 122. Cianciotto NP. Type II secretion and Legionella virulence. Curr Top Role of Three Type Ii Secreted Effectors in Legionella Pneumophila Microbiol Immunol 2013;376:81–102. Metabolism, Physiology and Pathogenesis. Evanston, IL.: North- 123. Mallama CA, McCoy-Simandle K, Cianciotto NP. The Type II western; 2011. Secretion System of Legionella pneumophila Dampens the MyD88 141. Pearce MM, Cianciotto NP. Legionella pneumophila secretes an and Toll-Like Receptor 2 Signaling Pathway in Infected Human endoglucanase that belongs to the family-5 of glycosyl hydro- Macrophages. Infect Immun 2017;85. lases and is dependent upon type II secretion. FEMS Microbiol Lett 124. McCoy-Simandle K, Stewart CR, Dao J, DebRoy S, Rossier O et al. 2009;300:256–264. Legionella pneumophila type II secretion dampens the cytokine 142. Herrmann V, Eidner A, Rydzewski K, Blädel I, Jules M et al. GamA response of infected macrophages and epithelia. Infect Immun is a eukaryotic-like glucoamylase responsible for glycogen- and 2011;79:1984–1997. starch-degrading activity of Legionella pneumophila. Int J Med 125. White RC, Cianciotto NP. Type II secretion is necessary for Microbiol 2011;301:133–139. optimal association of the Legionella-containing vacuole with 143. Rossier O, Dao J, Cianciotto NP. The type II secretion system macrophage Rab1B but enhances intracellular replication of Legionella pneumophila elaborates two aminopeptidases, mainly by Rab1B-Independent mechanisms. Infect Immun as well as a metalloprotease that contributes to differen- 2016;84:3313–3327. tial infection among protozoan hosts. Appl Environ Microbiol 126. Truchan HK, Christman HD, White RC, Rutledge NS, Cianci- 2008;74:753–761. otto NP. Type II Secretion Substrates of Legionella pneumophila 144. Zhang N, Yin S, Zhang W, Gong X, Zhang N et al. Crystal struc- Translocate Out of the Pathogen-Occupied Vacuole via a Semi- ture and biochemical characterization of an aminopepti- permeable Membrane. MBio 2017;8:e00870-17. dase LapB from Legionella pneumophila. J Agric Food Chem 2017;65:7569–7578. 127. Faucher SP, Mueller CA, Shuman HA. Legionella pneumophila transcriptome during intracellular multiplication in human 145. Vandersmissen L, De Buck E, Saels V, Coil DA, Anné J. A Legionella macrophages. Front Microbiol 2011;2:60. pneumophila collagen-like protein encoded by a gene with a vari- able number of tandem repeats is involved in the adherence and 128. Brüggemann H, Hagman A, Jules M, Sismeiro O, Dillies M-A et al. invasion of host cells. FEMS Microbiol Lett 2010;306:168–176. Virulence strategies for infecting phagocytes deduced from the in vivo transcriptional program of Legionella pneumophila. Cell 146. Abdel-Nour M, Duncan C, Prashar A, Rao C, Ginevra C et al. The Microbiol 2006;8:1228–1240. Legionella pneumophila collagen-like protein mediates sedimen- tation, autoaggregation, and pathogen- interactions. 129. Broich M, Rydzewski K, McNealy TL, Marre R, Flieger A. The Appl Environ Microbiol 2014;80:1441–1454. global regulatory proteins LetA and RpoS control phospholi- pase A, lysophospholipase A, acyltransferase, and other hydro- 147. Aragon V, Rossier O, Cianciotto NP. Legionella pneumophila genes lytic activities of Legionella pneumophila JR32. J Bacteriol that encode lipase and phospholipase C activities. Microbiology 2006;188:1218–1226. 2002;148:2223–2231. 130. Jules M, Buchrieser C. Legionella pneumophila adaptation to 148. Zusman T, Degtyar E, Segal G. Identification of a hypervariable intracellular life and the host response: clues from genomics and region containing new Legionella pneumophila Icm/Dot translo- transcriptomics. FEBS Lett 2007;581:2829–2838. cated substrates by using the conserved icmQ regulatory signa- ture. Infect Immun 2008;76:4581–4591. 131. DebRoy S, Dao J, Söderberg M, Rossier O, Cianciotto NP. Legionella pneumophila type II secretome reveals unique exopro- 149. Söderberg MA, Cianciotto NP. A Legionella pneumophila peptidyl- teins and a chitinase that promotes bacterial persistence in the prolyl cis-trans isomerase present in culture supernatants is lung. Proc Natl Acad Sci U S A 2006;103:19146–19151. necessary for optimal growth at low temperatures. Appl Environ Microbiol 2008;74:1634–1638. 132. Galka F, Wai SN, Kusch H, Engelmann S, Hecker M et al. Prot- eomic characterization of the whole secretome of Legionella 150. Aragon V, Kurtz S, Cianciotto NP. Legionella pneumophila major pneumophila and functional analysis of outer membrane vesi- acid phosphatase and its role in intracellular infection. Infect cles. Infect Immun 2008;76:1825–1836. Immun 2001;69:177–185. 133. Aurass P, Gerlach T, Becher D, Voigt B, Karste S et al. Life stage- 151. Dhatwalia R, Singh H, Reilly TJ, Tanner JJ. Crystal structure and specific proteomes ofLegionella pneumophila reveal a highly tartrate inhibition of Legionella pneumophila histidine acid phos- differential abundance of virulence-associated Dot/Icm effec- phatase. Arch Biochem Biophys 2015;585:32–38. tors. Mol Cell Proteomics 2016;15:177–200. 152. Gong X, Zhao X, Zhang W, Wang J, Chen X et al. Structural charac- 134. De Buck E, Höper D, Lammertyn E, Hecker M, Anné J. Differen- terization of the hypothetical protein Lpg2622, a new member of tial 2-D protein gel electrophoresis analysis of Legionella pneu- the C1 family peptidases from Legionella pneumophila. FEBS Lett mophila wild type and Tat secretion mutants. Int J Med Microbiol 2018;592:2798–2810. 2008;298:449–461. 153. Tyson JY. Novel Type II Secretion Substrates of Legionella pneu- 135. Rossier O, Cianciotto NP. The Legionella pneumophila tatB gene mophila Contribute to Infection of Multiple Amoebal Hosts. Evan- facilitates secretion of phospholipase C, growth under iron- ston, IL: Northwestern; 2014. limiting conditions, and intracellular infection. Infect Immun 154. Zhang N, Yin S, Liu S, Sun A, Zhou M et al. Crystal structure of 2005;73:2020–2032. lpg1832, a VirK family protein from Legionella pneumophila, 136. Debroy S, Aragon V, Kurtz S, Cianciotto NP. Legionella pneumophila reveals a novel fold for bacterial VirK proteins. FEBS Lett MIP, a surface-exposed peptidylproline cis-trans-isomerase, 2017;591:2929–2935. promotes the presence of phospholipase C-like activity in culture 155. Flieger A, Neumeister B, Cianciotto NP. Characterization of supernatants. Infect Immun 2006;74:5152–5160. the gene encoding the major secreted lysophospholipase A of

27 White and Cianciotto, Microbial Genomics 2019;5

Legionella pneumophila and its role in detoxification of lysophos- 176. O'Connor TJ, Boyd D, Dorer MS, Isberg RR. Aggravating genetic phatidylcholine. Infect Immun 2002;70:6094–6106. interactions allow a solution to redundancy in a bacterial path- 156. Creasey EA, Isberg RR. The protein SdhA maintains the integ- ogen. Science 2012;338:1440–1444. rity of the Legionella-containing vacuole. Proc Natl Acad Sci U S A 177. Ghosh S, O'Connor TJ. Beyond paralogs: the multiple layers of 2012;109:3481–3486. redundancy in bacterial pathogenesis. Front Cell Infect Microbiol 157. Banerji S, Bewersdorff M, Hermes B, Cianciotto NP, Flieger A. 2017;7:467. Characterization of the major secreted zinc metalloprotease- 178. Schroeder GN. The toolbox for uncovering the functions of dependent glycerophospholipid:cholesterol acyltransferase, Legionella Dot/Icm Type IVb secretion system effectors: current PlaC, of Legionella pneumophila. Infect Immun 2005;73:2899–2909. state and future directions. Front Cell Infect Microbiol 2017;7:528. 158. Rossier O, Dao J, Cianciotto NP. A type II secreted RNase of 179. Blander SJ, Szeto L, Shuman HA, Horwitz MA. An immuno- Legionella pneumophila facilitates optimal intracellular infection protective molecule, the major secretory protein of Legionella of hartmannella vermiformis. Microbiology 2009;155:882–890. pneumophila, is not a virulence factor in a guinea pig model of 159. Moffat JF, Edelstein PH, Regula DP, Cirillo JD, Tompkins LS. Legionnaires' disease. J Clin Invest 1990;86:817–824. Effects of an isogenic Zn-metalloprotease-deficient mutant of 180. James BW, Mauchline WS, Dennis PJ, Keevil CW. A study of iron Legionella pneumophila in a guinea-pig pneumonia model. Mol acquisition mechanisms of Legionella pneumophila grown in Microbiol 1994;12:693–705. chemostat culture. Curr Microbiol 1997;34:238–243. 160. Szeto L, Shuman HA. The Legionella pneumophila major secre- 181. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJE. The tory protein, a protease, is not required for intracellular growth Phyre2 web portal for protein modeling, prediction and analysis. or cell killing. Infect Immun 1990;58:2585–2592. Nat Protoc 2015;10:845–858. 161. Conlan JW, Williams A, Ashworth LAE. In vivo production of a 182. Liepinsh E, Généreux C, Dehareng D, Joris B, Otting G. NMR tissue-destructive protease by Legionella pneumophila in the structure of Citrobacter freundii AMPD, comparison with bacte- lungs of experimentally infected guinea-pigs. Microbiology riophage T7 lysozyme and homology with PGRP domains. J Mol 1988;134:143–149. Biol 2003;327:833–842. 162. Williams A, Baskerville A, Dowsett AB, Conlan JW. Immunocy- 183. Conesa A, Götz S, García-Gómez JM, Terol J, tochemical demonstration of the association between Legionella Talón M et al. Blast2GO: a universal tool for annotation, visu- pneumophila, its tissue-destructive protease, and pulmo- alization and analysis in functional genomics research. Bioinfor- nary lesions in experimental Legionnaires' disease. J Pathol matics 2005;21:3674–3676. 1987;153:257–264. 184. Tesh MJ, Morse SA, Miller RD. Intermediary metabolism in 163. Baskerville A, Conlan JW, Ashworth LA, Dowsett AB. Pulmonary Legionella pneumophila: utilization of amino acids and other damage caused by a protease from Legionella pneumophila. Br J compounds as energy sources. J Bacteriol 1983;154:1104–1109. Exp Pathol 1986;67:527–536. 185. Schunder E, Gillmaier N, Kutzner E, Eisenreich W, 164. Conlan JW, Baskerville A, Ashworth LAE. Separation of Legionella Herrmann V et al. Amino acid uptake and metabolism of Legionella pneumophila proteases and purification of a protease which pneumophila hosted by Acanthamoeba castellanii. J Biol Chem produces lesions like those of legionnaires" disease in guinea pig 2014;289:21040–21054. lung. Microbiology 1986;132:1565–1574. 186. Manske C, Hilbi H. Metabolism of the vacuolar pathogen 165. Berdal BP, Olsvik O, Myhre S, Omland T. Demonstration of Legionella and implications for virulence. Front Cell Infect Micro- extracellular chymotrypsin-like activity from various Legionella biol 2014;4:125. species. J Clin Microbiol 1982;16:452–457. 187. de Felipe KS, Pampou S, Jovanovic OS, Pericone CD, Ye SF 166. Thompson MR, Miller RD, Iglewski BH. In vitro production of an et al. Evidence for acquisition of Legionella type IV secretion extracellular protease by Legionella pneumophila. Infect Immun substrates via interdomain horizontal gene transfer. J Bacteriol 1981;34:299–302. 2005;187:7716–7726. 167. Berdal BP, Fossum K. Occurrence and immunogenicity of protein- 188. Nora T, Lomma M, Gomez-Valero L, Buchrieser C. Molecular ases from Legionella species. Eur J Clin Microbiol 1982;1:7–11. mimicry: an important virulence strategy employed by Legionella 168. Muller HE. Proteolytic action of Legionella pneumophila on human pneumophila to subvert host functions. Future Microbiol serum proteins. Infect Immun 1980;27:51–53. 2009;4:691–701. 169. Müller HE. Enzymatic profile of Legionella pneumophilia. J Clin 189. Lurie-Weinberger MN, Gomez-Valero L, Microbiol 1981;13:423–426. Merault N, Glöckner G, Buchrieser C et al. The origins of eukar- 170. Thorpe TC, Miller RD. Extracellular enzymes of Legionella pneu- yotic-like proteins in Legionella pneumophila. Int J Med Microbiol mophila. Infect Immun 1981;33:632–635. 2010;300:470–481. 171. Aragon V, Kurtz S, Flieger A, Neumeister B, Cianciotto NP. 190. Gomez-Valero L, Rusniok C, Cazalet C, Buchrieser C. Compara- Secreted enzymatic activities of wild-type and pilD-deficient tive and functional genomics of Legionella identified eukaryotic Legionella pneumophila. Infect Immun 2000;68:1855–1863. like proteins as key players in host?pathogen interactions. Front Microbiol 2011;2:Artn 208. 172. Flieger A, Gong S, Faigle M, Stevanovic S, Cianciotto NP et al. Novel lysophospholipase A secreted by Legionella pneumophila. 191. Horwitz MA. of the Legionnaires' disease bacte- Journal of Bacteriology 2001;183:2121–2124. rium (Legionella pneumophila) occurs by a novel mechanism: engulfment within a pseudopod coil. Cell 1984;36:27–33. 173. Aurass P, Schlegel M, Metwally O, Harding CR, Schroeder GN et al. The Legionella pneumophila Dot/Icm-secreted effector PlcC/ 192. Clemens DL, Lee BY, Horwitz MA. enters CegC1 together with PlcA and PlcB promotes virulence and macrophages via a novel process involving pseudopod loops. belongs to a novel zinc metallophospholipase C family present in Infect Immun 2005;73:5892–5902. bacteria and fungi. J Biol Chem 2013;288:11080–11092. 193. Celli J, Zahrt TC. Mechanisms of Francisella tularensis intracellular 174. Lang C, Hiller M, Flieger A. Disulfide loop cleavage of Legionella pathogenesis. Cold Spring Harb Perspect Med 2013;3:a010314. pneumophila PlaA boosts lysophospholipase A activity. Sci Rep 194. Abd H, Johansson T, Golovliov I, Sandström G, Forsman M. 2017;7:16313. Survival and growth of Francisella tularensis in Acanthamoeba 175. Lang C, Rastew E, Hermes B, Siegbrecht E, Ahrends R et al. Zinc castellanii. Appl Environ Microbiol 2003;69:600–606. metalloproteinase ProA directly activates Legionella pneumophila 195. Schrallhammer M, Castelli M, Petroni G. Phylogenetic relation- PlaC glycerophospholipid:cholesterol acyltransferase. J Biol ships among endosymbiotic R-body producer: bacteria providing Chem 2012;287:23464–23478. their host the killer trait. Syst Appl Microbiol 2018;41:213–220.

28 White and Cianciotto, Microbial Genomics 2019;5

196. Santos P, Pinhal I, Rainey FA, Empadinhas N, Costa J et al. 216. Haft DH, DiCuccio M, Badretdin A, Brover V, Chetvernin V et al. Gamma-Proteobacteria Aquicella lusitana gen. nov., sp. nov., and RefSeq: an update on prokaryotic genome annotation and cura- Aquicella siphonis sp. nov. infect protozoa and require activated tion. Nucleic Acids Res 2018;46:D851–D860. charcoal for growth in laboratory media. Appl Environ Microbiol 217. Joseph SJ, Cox D, Wolff B, Morrison SS, Kozak-Muiznieks NA 2003;69:6533–6540. et al. Dynamics of genome change among Legionella species. Sci 197. Gomez-Valero L, Rusniok C, Carson D, Mondino S, Pérez-Cobas AE Rep 2016;6:33442. et al. More than 18,000 effectors in the Legionella genus genome 218. Arslan-Aydoğdu EO, Kimiran A. An investigation of virulence provide multiple, independent combinations for replication in factors of Legionella pneumophila environmental isolates. Braz J human cells. Proc Natl Acad Sci U S A 2019;116:2265–2273. Microbiol 2018;49:189–199. 198. de Felipe KS, Glover RT, Charpentier X, Anderson OR, Reyes M 219. Fields BS, Barbaree JM, Sanden GN, Morrill WE. Virulence of a et al. Legionella eukaryotic-like type IV substrates interfere with strain associated with : an evalua- organelle trafficking. PLoS Pathog 2008;4:e1000117. tion using protozoan, cell culture, and guinea pig models. Infect 199. Matsuda S, Okada R, Tandhavanant S, Hiyoshi H, Gotoh K et al. Immun 1990;58:3139–3142. Export of a Vibrio parahaemolyticus toxin by the Sec and type III 220. Wadowsky RM, Wilson TM, Kapp NJ, West AJ, Kuchta JM et al. secretion machineries in tandem. Nat Microbiol 2019;4:781–788. Multiplication of Legionella spp. in tap water containing Hartman- 200. Degtyar E, Zusman T, Ehrlich M, Segal G. A Legionella effector nella vermiformis. Appl Environ Microbiol 1991;57:1950–1955. acquired from protozoa is involved in sphingolipids metabo- 221. Neumeister B, Schöniger S, Faigle M, Eichner M, lism and is targeted to the host cell mitochondria. Cell Microbiol Dietz K. Multiplication of differentLegionella species in Mono Mac 2009;11:1219–1235. 6 cells and in Acanthamoeba castellanii. Appl Environ Microbiol 201. Kubiak X, Dervins-Ravault D, Pluvinage B, Chaffotte AF, 1997;63:1219–1224. Gomez-Valero L et al. Characterization of an acetyltransferase 222. Edelstein PH, Edelstein MA, Shephard LJ, Ward KW, Ratcliff RM. that detoxifies aromatic chemicals in Legionella pneumophila. sp. nov., isolated from human respiratory spec- Biochem J 2012;445:219–228. imens in California, USA, and South Australia. Int J Syst Evol 202. Abrahão J, Silva L, Silva LS, Khalil JYB, Rodrigues R et al. Tailed Microbiol 2012;62:1766–1771. giant Tupanvirus possesses the most complete translational 223. Adeleke AA, Fields BS, Benson RF, Daneshvar MI, Pruckler JM apparatus of the known virosphere. Nat Commun 2018;9:749. et al. sp. nov., sp. 203. Gomez-Valero L, Buchrieser C. Genome dynamics in Legionella: nov. and sp. nov.: three unusual new Legionella the basis of versatility and adaptation to intracellular replication. species. Int J Syst Evol Microbiol 2001;51:1151–1160. Cold Spring Harb Perspect Med 2013;3:a009993. 224. Cazalet C, Gomez-Valero L, Rusniok C, Lomma M, Dervins- 204. Moliner C, Raoult D, Fournier PE. Evidence that the intra-amoebal Ravault D et al. Analysis of the genome acquired a sterol reductase gene from and transcriptome uncovers unique strategies to cause eukaryotes. BMC Res Notes 2009;2:51. Legionnaires' disease. PLoS Genet 2010;6:e1000851. 205. Hingamp P, Grimsley N, Acinas SG, Clerissi C, Subirana L et al. 225. Bacigalupe R, Lindsay D, Edwards G, Fitzgerald JR. Popu- Exploring nucleo-cytoplasmic large DNA viruses in tara Oceans lation genomics of Legionella longbeachae and hidden microbial metagenomes. Isme J 2013;7:1678–1695. complexities of infection source attribution. Emerg Infect Dis 206. Yu Z, An B, Ramshaw JAM, Brodsky B. Bacterial collagen- 2017;23:750–757. like proteins that form triple-helical structures. J Struct Biol 226. Kozak NA, Buss M, Lucas CE, Frace M, Govil D et al. Virulence 2014;186:451–461. factors encoded by Legionella longbeachae identified on the basis 207. Stebbins CE, Galán JE. Structural mimicry in bacterial virulence. of the genome sequence analysis of clinical isolate D-4968. J Nature 2001;412:701–705. Bacteriol 2010;192:1030–1044. 208. Ozanic M, Gobin I, Brezovec M, Marecic V, Trobonjaca Z et al. F. 227. Cho H, Cronan JE. Escherichia coli thioesterase I, molecular novicida-Infected A. castellanii Does Not Enhance Bacterial Viru- cloning and sequencing of the structural gene and identification lence in Mice. Front Cell Infect Microbiol 2016;6:56. as a periplasmic enzyme. J Biol Chem 1993;268:9238–9245. 209. Ozanic M, Marecic V, Abu Kwaik Y, Santic M. The divergent intra- 228. Burstein D, Amaro F, Zusman T, Lifshitz Z, Cohen O et al. Genomic cellular lifestyle of Francisella tularensis in evolutionarily distinct analysis of 38 Legionella species identifies large and diverse host cells. PLoS Pathog 2015;11:e1005208. effector repertoires. Nat Genet 2016;48:167–175. 210. Al-Khodor S, Kalachikov S, Morozova I, Price CT, Abu Kwaik Y. The 229. Williams KP. Integration sites for genetic elements in prokary- PmrA/PmrB two-component system of Legionella pneumophila otic tRNA and tmRNA genes: sublocation preference of integrase is a global regulator required for intracellular replication within subfamilies. Nucleic Acids Res 2002;30:866–875. macrophages and protozoa. Infect Immun 2009;77:374–386. 230. Veltri D, Wight MM, Crouch JA. SimpleSynteny: a web-based tool 211. Sahr T, Rusniok C, Impens F, Oliva G, Sismeiro O et al. The for visualization of microsynteny across multiple species. Nucleic Legionella pneumophila genome evolved to accommodate Acids Res 2016;44:W41–W45. multiple regulatory mechanisms controlled by the CsrA-system. 231. Tan Y, Luo Z-Q. Legionella pneumophila SidD is a deAMPylase that PLoS Genet 2017;13:e1006629. modifies Rab1. Nature 2011;475:506–509. 212. Tanner JR, Li L, Faucher SP, Brassinga AKC. The CpxRA two- 232. Tan Y, Arnold RJ, Luo ZQ. Legionella pneumophila regulates the component system contributes to Legionella pneumophila viru- small GTPase Rab1 activity by reversible phosphorylcholination. lence. Mol Microbiol 2016;100:1017–1038. Proc Natl Acad Sci U S A 2011;108:21212–21217. 213. Costa J, d'Avó AF, da Costa MS, Veríssimo A. Molecular evolu- 233. Flieger A, Gong S, Faigle M, Northoff H, Neumeister B. In vitro tion of key genes for type II secretion in Legionella pneumophila. secretion kinetics of proteins from Legionella pneumophila in Environ Microbiol 2012;14:2017–2033. comparison to proteins from non-pneumophila species. Microbi- 214. Qin T, Zhou H, Ren H, Liu W. Distribution of secretion systems in ology 2001;147:3127–3134. the genus Legionella and Its correlation with pathogenicity. Front 234. McIntyre M, Quinn FD, Fields PI, Berdal BP. Rapid identification of Microbiol 2017;8:388. Legionella pneumophila zinc metalloprotease using chromogenic 215. Gomez-Valero L, Rusniok C, Rolando M, Neou M, Dervins- detection. APMIS 1991;99:316–320. Ravault D et al. Comparative analyses of Legionella species iden- 235. Flieger A, Gong S, Faigle M, Deeg M, Bartmann P et al. Novel tifies genetic features of strains causing Legionnaires' disease. phospholipase A activity secreted by Legionella species. J Bacte- Genome Biol 2014;15:505. riol 2000;182:1321–1327.

29 White and Cianciotto, Microbial Genomics 2019;5

236. Baine WB. Cytolytic and phospholipase C activity in Legionella 240. van Schaik EJ, Chen C, Mertens K, Weber MM, Samuel JE. Molec- species. J Gen Microbiol 1985;131:1383–1391. ular pathogenesis of the obligate intracellular bacterium . Nat Rev Microbiol 2013;11:561–573. 237. Nagai H, Kubori T. Type IVB secretion systems of Legionella and 241. Mediannikov O, Sekeyová Z, Birg ML, Raoult D. A novel obligate other gram-negative bacteria. Front Microbiol 2011;2:136. intracellular gamma-proteobacterium associated with ixodid 238. Gong J, Qing Y, Zou S, Fu R, Su L et al. Protist-Bacteria associa- ticks, Diplorickettsia massiliensis, Gen. Nov., Sp. Nov. PLoS One tions: gammaproteobacteria and alphaproteobacteria are prev- 2010;5:e11478. alent as digestion-resistant bacteria in ciliated protozoa. Front 242. Fields BS, Benson RF, Besser RE, Legionella BRE. Legionella and Microbiol 2016;7:498. Legionnaires' disease: 25 years of investigation. Clin Microbiol Rev 2002;15:506–526. 239. Cordaux R, Paces-Fessy M, Raimond M, Michel- 243. Mehari YT, Jason Hayes B, Redding KS, Mariappan PVG, Salzat A, Zimmer M et al. Molecular characterization and evolu- Gunderson JH et al. Description of 'Candidatus Berkiella aquae' and tion of arthropod-pathogenic Rickettsiella bacteria. Appl Environ 'Candidatus Berkiella cookevillensis', two intranuclear bacteria of Microbiol 2007;73:5045–5047. freshwater amoebae. Int J Syst Evol Microbiol 2016;66:536–541.

Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review – average time to first decision is 4–6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 80% of our authors rate our submission process as ‘excellent’ or ‘very good’. 5. Your article will be published on an interactive journal platform with advanced metrics.

Find out more and submit your article at microbiologyresearch.org.

30

Top View